Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area

Transcription

1 CAPCOG FY14-15 PGA FY14-1 Deliverable Amendment 1 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area Prepared by the Capital Area Council of Governments October 8, 2015 PREPARED UNDER A GRANT FROM THE TEXAS COMMISSION ON ENVIRONMENTAL QUALITY The preparation of this report was financed through grants from the State of Texas through the Texas Commission on Environmental Quality. The content, findings, opinions, and conclusions are the work of the author(s) and do not necessarily represent findings, opinions, or conclusions of the TCEQ.

2 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area Executive Summary This ozone conceptual model describes and analyzes the conditions that lead to high ozone concentrations within the Austin-Round Rock Metropolitan Statistical Area (MSA), which consists of Bastrop, Caldwell, Hays, Travis, and Williamson Counties. This ozone conceptual model expands the previous conceptual model for the Austin-Round Rock MSA developed by the University of Texas at Austin (UT) in 2013, which used data from , by incorporating data collected during and adding additional types of analysis. 1 For this report, five types of high ozone days were analyzed: 1. Peak 8-hour ozone concentrations over 75 parts per billion (ppb); 2. Peak 8-hour ozone concentrations over 70 ppb; 3. Peak 8-hour ozone concentrations over 65 ppb; 4. The top four eight-hour ozone concentrations in a given year; 5. The top ten eighth-hour ozone concentrations in a given year. This conceptual model describes the annual, seasonal, and day-of-week variability of high ozone concentrations, defined as exceedances of the current standard and low and high ends of the range proposed by Environmental Protection Agency (EPA) for a new ozone standard, and includes a description of the large-scale weather patterns and associated local meteorological conditions typically experienced during high ozone episodes in the Austin- Round Rock MSA. This report includes an analysis of the local, intra-state, and inter-state geographic areas often upwind of Central Texas prior to high ozone days in the Austin- Round Rock MSA (which correspond to potential source regions of background ozone entering the Austin area), and provides estimates of background and locally-formed ozone using available monitoring data. Section 1 provides an introduction and review of important background information. Section 2 describes the data sources used for the analyses in this report. Section 3 includes general analysis of ozone trends, geographic variation in peak ozone levels within the region, and the frequency of high ozone conditions within the region by year, month, day of the week, and time of day. Section 4 includes an analysis of typical meteorological conditions on high ozone days, including analysis of temperature, wind speed, relative humidity, and wind direction. Section 5 includes an analysis of background and locally-formed ozone levels within the region based on the lowest and highest peak 8-hour ozone concentrations monitored within the region. Section 6 includes an analysis of regional transport patterns prior to high ozone days. Section 7 provides an analysis of large-scale weather patterns during high ozone episodes, including commonly observed large-scale weather features and typical upwind source regions. Section 8 includes analyses of meteorological conditions during three periods of high ozone that are currently being used by TCEQ and EPA for ozone modeling, but which CAPCOG has not started to use yet (Fall 2006, Fall 2011, and June 2012). Section 9 includes an analysis of the relationship between large-scale transport conditions and the Austin-Round Rock MSA maximum ozone concentration. A quality-assurance (QA) report is included as an appendix. Questions Posed in 2014 Draft Modeling Guidance Section 2.1 of EPA s draft December 2014 update 2 to ozone modeling guidance, Guidance on the Use of Models and Other Analyses to Demonstrating Attainment of Air Quality goals for Ozone, 1 Austin_Area_Conceptual_Model_2012.pdf 2 Page 2 of 211

3 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area PM2.5, and Regional Haze 3 provides some general questions to use to develop a conceptual model. The answers to these questions derived from this conceptual model are provided below. What are the pollutants of concern for the area? Ground-level ozone is the primary pollutant of concern for the Austin-Round Rock MSA, given that the region s ozone levels are much closer to the level of the National Ambient Air Quality Standard (NAAQS) for ozone than the region s air pollution concentrations are relative to any other NAAQS. What is the attainment/nonattainment status of the area? All of the counties in the Austin-Round Rock MSA are designated attainment/unclassifiable for all NAAQS. What is the geographic scope of poor air quality? For this study, poor air quality is defined using three thresholds: over 75 parts per billion, over 70 parts per billion, and over 65 parts per billion. Over the period analyzed, monitors in Bastrop, Hays, Travis, and Williamson Counties have all measured at least one eight-hour ozone concentration above 75 ppb, with each county having a monitor with a measurement above 75 ppb as recently as There has only been an ozone monitor in Caldwell County since 2013, and it has not yet measured 8-hour ozone concentrations above 75 ppb, but it has measured one 8-hour ozone concentrations above 70 ppb during this time frame. What is the temporal scope of poor air quality? Eight-hour ozone levels above 65 ppb can occur as early as the beginning of March and as late as the beginning of November. Levels above 70 ppb can occur as early as mid-march and as late as the end of October. Levels above 75 ppb only occur between May and early October. High ozone levels most frequently occur from Mid-May to late June and from mid-august to early October. They occur least frequently on Sundays, and most frequently on Fridays and Saturdays. Typically, peak 8-hour ozone concentrations occur between 10 am and 6 pm or between 11 am and 7 pm. What are the air quality trends in the area? Ozone levels are improving. From , the region s ozone design value has declined from 82 ppb to 69 ppb, a 16% decrease. From , the region s ozone design value has decreased an average of 1.3 ppb per year. What are the suspected mechanisms for formation of poor air quality levels? High peak daily temperatures, large changes in temperature in a given day, low relative humidity, slow winds, and wind directions traveling across continental areas are the typical conditions that are conducive to high ozone formation within the region. What are the sources of emissions that may contribute to poor air quality? Of the portion of peak ozone levels in the Austin-Round Rock MSA attributable to anthropogenic emissions in the U.S., about a third of the ozone impact comes from emissions within the MSA, a third comes from emissions elsewhere in Texas, and a third comes from emissions elsewhere in the country. As anthropogenic emissions in the U.S. decrease, the relative importance of policyrelevant background ozone levels, biogenic emissions, and wildfires will increase. Within the MSA, anthropogenic NOX emissions have about times the impact on peak ozone levels as 3 Page 3 of 211

4 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area the same quantity of emissions of anthropogenic VOC. On-road sources accounted for about 60% of local NOX emissions in the 2011 National Emissions Inventory (NEI), with non-road sources accounting for 20%, and stationary sources accounting for another 20%, not counting the abnormally high wildfire emissions that occurred in Are there unique meteorological influences on local air quality levels? Some of the worst high-ozone days can occur when there are winds that come out of the west/northwest early in the morning and then change direction mid-morning to southsoutheast through east-southeast. This phenomenon creates a situation in which the urban area s emissions can wind up doubling up the impact on peak ozone levels as the same air passes back over the area. Regional Ozone Overview The National Ambient Air Quality Standard (NAAQS) that the Austin-Round Rock MSA is closest to violating is the primary NAAQS for ozone. While the current standard level is ppm (75 ppb) averaged over eight hours, the EPA has proposed setting a new standard between ppb. The final standard will be announced by October 1, The Austin-Round Rock MSA design value has been tracking downward since 1999, and is in attainment of the current standard; its most recent design value for was 69 ppb. When and Where High Ozone Days Occur Ozone concentrations in the Austin-Round Rock MSA have been steadily decreasing from While there were 50 days over 65 ppb in 2006, there were only 6 in And whereas there were 17 days when ozone levels were over 75 ppb, there were none in Most high ozone days during occurred during April-October, with a strong peak in May-June, and again in August-September, though the seasonal distribution of high ozone varied year to year. There was less of a pattern for ozone formation by the days of the week, but CAPCOG s analysis shows that Sunday has the fewest number of high ozone days overall, and generally rises through the week into Friday and Saturday. Finally, the peak ozone concentrations (1-hour measurements) in a given high-ozone day occurred most frequently between 12pm and 5pm. Local Surface Meteorological Conditions on High Ozone Days Using all available meteorological data, CAPCOG analyzed temperature, relative humidity, wind speeds and wind directions to determine the conditions conducive to higher ozone formation. At the two regulatory monitors, continuous air monitoring station (CAMS) 3 and CAMS 38, high ozone days had daily temperature changes of greater than 23 degrees Fahrenheit, wind speeds of less than 8.3 miles per hour, relative humidity of less than 38%, and wind directions from north-northeast (NNE) clockwise to South-Southeast (SSE). The figure below summarizes the typical conditions for each station above each threshold level analyzed. Page 4 of 211

5 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area Table ES- 1. Typical Meteorological Conditions Conducive to Ozone Formation for CAMS 3 & 38, Monitor CAMS 3 CAMS 38 8-Hour Average Ozone Concentration Daily Temperature Change Page 5 of 211 Wind Speeds 2pm Relative Humidity Wind Direction (Clockwise) >75 ppb >24.9 o F <5.4 mph <38.0% NNE; E-SSE >70 ppb >23.8 o F <6.1 mph <39.9% NNE; SSE >65 ppb >23.0 o F <8.3 mph <43.1% NNE; SE-SSE >75 ppb >24.1 o F <4.8 mph <34.1% E >70 ppb >24.4 o F <5.5 mph <39.1% E-S >65 ppb >23.3 o F <7.4 mph <43.3% ENE-S Estimates of Background and Locally-Formed Ozone Using the available local regulatory and CAPCOG monitoring data from within the MSA to determine the minimum and maximum peak 8-hour ozone concentrations, CAPCOG estimated that background ozone represented the upwind concentrations were 71-88% of the downwind concentrations in the area. The results for each analysis are presented in the figure below. These results are consistent with modeling analyses showing a similar level of local contribution to peak 8-hour ozone concentrations. Table ES- 2. Background Ozone Concentrations in Austin-Round Rock MSA monitoring system, Ozone Level Average Background Concentration Average Local Contribution Average Background as % of Max >75 ppb 62.5 ppb 18.3 ppb 77.4% >70 ppb 59.6 ppb 16.5 ppb 78.1% >65 ppb 56.9 ppb 15.2 ppb 80.4% Ozone Transport Back-trajectories maps were developed for regulatory monitors, CAMS 3 and 38, to visually summarize the most frequent geographic areas upwind of the Austin-Round Rock MSA prior to days with high ozone concentrations. The upwind geographic areas identified by the 48-hour back-trajectories correspond to the potential source regions of background ozone entering the area. Maps were generated to investigate the spatial variability for both the early (April-July) and late (August-October) ozone seasons. From this analysis, it was determined that: The most common area located upwind of Central Texas prior to Austin-Round Rock MSA high ozone events included the Central Plains, the Mississippi and Ohio River Valleys, and (less commonly) the Southeastern U.S. Within Texas, upwind areas are variable in the early season (April-July), including the north, northeast, and southeast directions. In the late part of the season (August-October) backtrajectories suggest transport from the northeast and east. A relative lack of back-trajectories originating from maritime regions suggests the inflow of continental air into Central Texas may be a necessary condition for high ozone in the area. Large-Scale Weather Patterns The conditions conducive to the transport, formation, and accumulation of ozone in Central Texas are primarily dependent on the prevailing large-scale weather patterns. High ozone episodes are usually preceded by the passage of a cold front through Central Texas. Following the front, a surface ridge of high pressure often extends south into Texas. While some cold

6 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area fronts were accompanied by strong gusty winds and the transport of noticeably cooler air into Texas, other cold fronts primarily represented a transition zone into drier continental air, which is typically associated with clear skies, warm temperatures, and light wind speeds at the surface. Monitor-specific observations are as follows: CAMS 3: Episodes in which CAMS 3 measured the highest ozone in the region generally had more stagnant and variable winds than other monitors, and 10 of the 13 days over 75ppb were between April and July, with relatively shorter ranges for 5-day back trajectories. Unlike the rest of the monitors, no long-range transport came from Canada or the northeastern US. CAMS 38: There were only four days in which CAMS 38 measured the highest ozone in the region over 75ppb, and three of those had long-range wind patterns from the northeastern US, and occurred between May and June. These were consistent with the general meteorological conditions following a cold front. CAMS 614: Although the Dripping Springs monitor also had predominantly northeasterly winds on high ozone days, the majority of these episodes occurred between August and October. Back-trajectories indicate a constant wind direction in the lower atmosphere, and are consistent with the transport of the urban plume over the monitor, as well as the cold front passage conditions. CAMS 674, 690, & 6602: These three monitors located in the northern part of the MSA showed wind patterns indicating recirculation of air over the urban area. The episodes that did not follow this pattern showed transport from the north and east. CAMS 675, 684, & 1675: Three-quarters of the episodes examined showed back-trajectories consistent with northerly winds entering Central Texas from the lower troposphere. The remaining 25% demonstrated transport from the south, San Antonio and further afield. Meteorological Conditions During New Photochemical Modeling Episodes Three different periods of high ozone within the Austin-Round Rock MSA were examined by AACOG as part of this analysis for future use by CAPCOG: Fall 2006, Fall 2011, and June Fall 2006 (August-October): A surface ridge of high pressure, associated with clear skies, warm temperatures, and light winds, extended south or southwest into Central Texas. Episodes typically were initiated by the passage of a cold front into or through Central Texas, transporting continental air, and high ozone followed the front one to two days later. Fall 2011 (August-October): Meteorological conditions during this period were extreme and not straightforward. The summer preceding it was characterized by all-time record drought and heat, but in late August a predominant weather pattern changed, transporting continental air into Texas. Following that period, more episodes were associated with wind shifts and recirculation of pollutants over Central Texas, and finally, Tropical Storm Lee, which influenced pressure systems to create periods of higher ozone in the area. June 2012: Two high ozone episodes occurred during this month. The first was initiated by the passage of a cold front that transported drier continental air into Central Texas, and resulted in several days of high ozone. The second event was initiated by the close approach of a frontal boundary, which created a week-long period of stagnant air flow, with some days experiencing 180 degree wind shifts. Page 6 of 211

7 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area Table of Contents Executive Summary... 2 Questions Posed in 2014 Draft Modeling Guidance... 2 What are the pollutants of concern for the area?... 3 What is the attainment/nonattainment status of the area?... 3 What is the geographic scope of poor air quality?... 3 What is the temporal scope of poor air quality?... 3 What are the air quality trends in the area?... 3 What are the suspected mechanisms for formation of poor air quality levels?... 3 What are the sources of emissions that may contribute to poor air quality?... 3 Are there unique meteorological influences on local air quality levels?... 4 Regional Ozone Overview... 4 When and Where High Ozone Days Occur... 4 Local Surface Meteorological Conditions on High Ozone Days... 4 Estimates of Background and Locally-Formed Ozone... 5 Ozone Transport... 5 Large-Scale Weather Patterns... 5 Meteorological Conditions During New Photochemical Modeling Episodes... 6 Table of Contents Introduction Conceptual Model Definition Austin-Round Rock MSA Ozone Levels Relationships Between Emissions and Air Quality Key Emission Sources or Source Categories Comparison of Emission Trends for Annual and/or Seasonal/Episodic Periods to Corresponding Air Quality Trends Resources on Other Topics on Relationship between Emissions and Air Quality Local Air Quality Planning Relevance of Conceptual Model for Future Photochemical Modeling Data Used in This Conceptual Model Regional Ozone Monitoring Data National Weather Service Synoptic Analyses HYSPLIT Back-Trajectories Temporal and Spatial Patterns in High Ozone Levels in the Region Page 7 of 211

8 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area 3.1 Location of Daily Peak Ozone Concentrations Fourth Highest Eight-Hour Ozone Concentrations Frequency of High Ozone Days by Year Frequency of High Ozone Days by Month Frequency of High Ozone Days by Day of the Week Frequency of High Ozone Days by Time of Day Local Meteorological Conditions Temperature Wind Speed Relative Humidity Wind Direction Background and Locally-Formed Ozone Regional Transport Patterns Prior to High Ozone Days Regional Transport at CAMS Regional Transport at CAMS Large-Scale Weather Patterns during High Ozone Episodes Commonly Observed Large-Scale Weather Features Upwind Geographic Source Regions Meteorological Conditions During Ozone Episodes Currently Used in Selected TCEQ and EPA Modeling Platforms August October August October June Relationship between Large-Scale Transport Conditions and the Austin Maximum Ozone 201 Appendix A: Quality Assurance Memo Table ES- 1. Typical Meteorological Conditions Conducive to Ozone Formation for CAMS 3 & 38, Table ES- 2. Background Ozone Concentrations in Austin-Round Rock MSA monitoring system, Table Annual Emissions by Tier 1 Source Category (tons per year) Table 1-2. Comparison of 2006 and 2012 photochemical modeling NOX emissions inventories for the MSA Table 1-3. Comparison of MSA, Texas, and U.S. NO X emissions in 2008 and 2011 (minus miscellaneous ) Table 2-1. Description and location of ozone monitoring stations in the Austin-Round Rock MSA Page 8 of 211

9 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area Table th Highest daily maximum 8-hour ozone concentrations (ppb) for Austin-Round Rock MSA Monitors, Table 4-1. Comparison of 2012 and 2015 Conceptual Model Meteorological Analyses Table 4-2. Ozone-maximum temperature correlation for all Austin-Round Rock MSA monitors, Table th and 100 th Percentile for maximum daily temperature for CAMS 3 & CAMS 38, Table th and 100 th Percentile for diurnal temperature change for CAMS 3 & CAMS 38, Table th Percentile for wind speed for CAMS 3 & CAMS 38, Table 4-6. Ozone-wind speed correlation for all Austin-Round Rock MSA monitors, Table 4-7. Maximum wind speeds for CAMS 3 & CAMS 38, Table 4-8. Ozone-relative humidity correlation for all Austin-Round Rock MSA monitors, Table th Percentile for Average Daily Relative Humidity and 2 pm Relative Humidity for CAMS Table th Percentile for Average Daily Relative Humidity and 2 pm Relative Humidity for CAMS Table th Percentile for Average Daily Relative Humidity and 2 pm Relative Humidity for CAMS Table th Percentile for Average Daily Relative Humidity and 2 pm Relative Humidity for CAMS Table th Percentile for Average Daily Relative Humidity and 2pm Relative Humidity for CAMS Table th Percentile for Average Daily Relative Humidity and 2pm Relative Humidity for CAMS Table 5-1. Monthly and total average statistics on >70ppb days based on the Austin-Round Rock MSA minimum and maximum ozone concentrations, Table 6-1. Bin counts for hourly 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 3, Table 6-2. Bin counts for hourly 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 3 for April-July ( ) Table 6-3. Bin counts for hourly 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 3 for August-October ( ) Table 6-4. Bin counts for hourly 1000m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 3, Table 6-5. Bin counts for hourly 1000m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 3 for April-July ( ) Table 6-6. Bin counts for hourly 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 3 for August-October ( ) Table 6-7. Bin counts for hourly 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 38, Table 6-8. Bin counts for hourly 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 38 for April-July ( ) Table 6-9. Bin counts for hourly 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 38 for August-October ( ) Table Bin counts for hourly 1000m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 38, Page 9 of 211

10 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area Table Bin counts for hourly 1000m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 38 for April-July ( ) Table Bin counts for hourly 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 38 for August-October ( ) Table high ozone episodes reviewed for the Austin area Table 7-2. Local meteorological measurements at Austin Northwest and large-scale weather features for high ozone episodes in Austin during Table 8-1. Daily local meteorological measurements at Austin Northwest and large-scale weather features for August October Table 8-2. Daily local meteorological measurements at Austin Northwest and large-scale weather features for August October Table 8-3. Daily local meteorological measurements at Austin Northwest and large-scale weather features for June Figure 1-1. Ozone Design Value Trend for the Austin-Round Rock MSA, Figure 2-1. Map of ozone CAMS in the Austin-Round Rock MSA active between Figure 3-1. Number of area-wide maximum ozone concentrations measured at each ozone monitoring station Figure 3-2. Likelihood that a station could measure the highest 8-hour ozone average on a high ozone day, Figure th highest 8-hour ozone concentrations at Austin-Round Rock MSA CAMS Figure 3-4. Annual number of high ozone days at one or more Austin area monitor, Figure 3-5. Maximum 8-hour ozone concentration >75ppb measured at regulatory and CAPCOG monitors, Figure 3-6. Maximum 8-hour ozone concentration >70ppb measured at regulatory and CAPCOG monitors, Figure 3-7. Maximum 8-hour ozone concentration >65ppb measured at regulatory and CAPCOG monitors, Figure 3-8. Monthly distribution of high ozone days for all monitoring sites, Figure 3-9. Annual number of early and late season days > 70 ppb, Figure Monthly distribution of highest four ozone days for CAMS 3 & CAMS 38, Figure Monthly distribution of highest ten ozone days for CAMS 3 & CAMS 38, Figure Number of high ozone days by day of the week at all Austin area monitors, Figure Number of high ozone measurements by day of week at regulatory monitors CAMS 3 & 38 only, Figure Day of the week distribution of highest four ozone days for CAMS 3 & CAMS 38, Figure Day of the week distribution of highest ten ozone days for CAMS 3 & CAMS 38, Figure Distribution of peak 1-hour ozone concentrations by hour for all Austin-Round Rock MSA monitors, Figure Hourly distribution of peak 1-hour ozone for highest four 8-hour concentrations, CAMS 3 & CAMS 38, Page 10 of 211

11 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area Figure Hourly distribution of peak 1-hour ozone for highest ten 8-hour concentrations, CAMS 3 & CAMS 38, Figure Distribution of times for peak 8-hour ozone concentrations for all Austin-Round Rock MSA monitors, Figure Distribution of times for highest four 8-hour ozone concentrations, CAMS 3 and CAMS 38, Figure Distribution of times for highest ten 8-hour ozone concentrations, CAMS 3 and CAMS 38, Figure 4-1. Maximum 8-hour ozone concentrations at CAMS 3 (ppb) vs. maximum daily temperature ( o F) at ABIA Figure 4-2. Maximum 8-hour ozone concentration at CAMS 3 (ppb) vs. diurnal differential temperature ( o F) at ABIA Figure 4-3. Maximum 8-hour ozone concentrations (ppb) vs. average wind speed (mph) between 6am-6pm at CAMS 3 (ppb), Figure 4-4. Maximum ozone concentrations (ppb) at CAMS 3 vs. daily average relative humidity (%) at ABIA, Figure 4-5. Maximum ozone concentrations (ppb) at CAMS 3 vs. daily 2pm relative humidity (%) at ABIA, Figure 4-6. Resultant wind direction 6am 6 pm on high ozone days for CAMS 3, 38, 614, 674, 690, and , Figure 4-7. Resultant wind direction 6am 6 pm on high ozone days at CAMS 613, 684, 6602, 1604, Liberty Hill, and Elroy, Figure 4-8. Comparison of wind direction on days with ozone >75ppb to days with ozone <=75ppb, controlling for meteorological conditions at CAMS Figure 4-9. Comparison of wind direction on days with ozone >70 ppb to days with ozone <=70ppb, controlling for meteorological conditions at CAMS Figure Comparison of wind direction on days with ozone >65ppb to days with ozone <=65ppb, controlling for meteorological conditions at CAMS Figure Comparison of wind direction on days with ozone >75ppb to days with ozone <=75ppb, controlling for meteorological conditions at CAMS Figure Comparison of wind direction on days with ozone >70ppb to days with ozone <=70ppb, controlling for meteorological conditions at CAMS Figure 4-13 Comparison of wind direction on days with ozone >65ppb to days with ozone <=65ppb, controlling for meteorological conditions at CAMS Figure Comparison of wind direction on high ozone days to low ozone days at CAMS Figure Comparison of wind direction on high ozone days to low ozone days at CAMS Figure 5-1. Average minimum/maximum ozone concentrations for different high ozone levels in the Austin-Round Rock MSA, , compared to analysis Figure 5-2. Average minimum/maximum ozone concentrations for different high ozone levels in the Austin-Round Rock MSA, compared to lower ozone days, Figure m HYSPLIT 48-hour back trajectories originating at CAMS 3 (Austin Northwest), Figure 6-2. (a) 50m HYSPLIT 48-hour back trajectories originating at CAMS 3 during April-July ( ); (b) 50m HYSPLIT 48-hour back trajectories originating at CAMS 3 during August- October ( ) Figure m HYSPLIT 48-hour back trajectories originating at CAMS 3, Page 11 of 211

12 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area Figure 6-4. (a) 1000m HYSPLIT 48-hour back trajectories originating at CAMS 3 during April-July ( ); (b) 1000m HYSPLIT 48-hour back trajectories originating at CAMS 3 during August- October ( ) Figure m HYSPLIT 48-hour back trajectories originating at CAMS 3, Figure 6-6. (a) 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 3 during April-July ( ); (b) 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 3 during August-October ( ) Figure m HYSPLIT 48-hour back trajectory hourly points ending at CAMS 3, Figure 6-8. (a) 1000m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 3 during April-July ( ); (b) 1000m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 3 during August-October ( ) Figure m HYSPLIT 48-hour back trajectories originating at CAMS 38, Figure (a) 50m HYSPLIT 48-hour back trajectories originating at CAMS 38 during April-July ; (b) August-October, Figure m HYSPLIT 48-hour back trajectories originating at CAMS 38, Figure m HYSPLIT 48-hour back trajectories originating at CAMS 38 during April-July ; (b) August-October, Figure m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 38, Figure (a) 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 3 during April-July ( ); (b) 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 38 during August-October ( ) Figure m HYSPLIT 48-hour back trajectory hourly points ending at CAMS 38, Figure (a) 1000m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 38 during April-July ( ); (b) 1000m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 38 during August-October ( ) Figure 7-1. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on May 14, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure 7-2. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on May 16, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure 7-3. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on May 17, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure 7-4. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on May 21, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure 7-5. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on June 1, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure 7-6. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on June 25, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Page 12 of 211

13 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area Figure 7-7. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on June 26, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure 7-8. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on June 27, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure 7-9. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 10, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 11, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 20, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 21, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 10, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on May 13, 2013 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on June 3, 2013 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on June 4, 2013 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on July 3, 2013 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on July 4, 2013 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on July 5, 2013 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 17, 2013 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 18, 2013 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 23, 2013 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Page 13 of 211

14 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 24, 2013 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 25, 2013 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on May 16, 2014 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure 8-1. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 30, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure 8-2. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 31, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure 8-3. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 1, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure 8-4. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 2, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure 8-5. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 3, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure 8-6. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 4, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure 8-7. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 6, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure 8-8. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 7, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure 8-9. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 8, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 14, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 20, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 26, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 27, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Page 14 of 211

15 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on October 5, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on October 6, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on October 8, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on October 9, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 26, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 27, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 28, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 29, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 4, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 6, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 7, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 8, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 9, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 10, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 11, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 12, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Page 15 of 211

16 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 13, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 20, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 22, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 24, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on October 2, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on October 3, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km Figure 9-1. Number of days when a monitor measured the maximum region-wide 8-hour ozone maximum concentration when at least one measurement was >=75 ppb Figure 9-2. Inter-state back-trajectories (based on 5-day HYSPLIT back-trajectories initiated at 1 km AGL) on 13 days during that had the maximum Austin concentration at CAMS 3 (Austin Northwest) Figure 9-3. Inter-state back-trajectories (based on 5-day HYSPLIT back-trajectories initiated at 1 km AGL) on 14 days during that had the maximum Austin concentration at CAMS 38 (Audubon) Figure 9-4. Inter-state back-trajectories (based on 5-day HYSPLIT back-trajectories initiated at 1 km AGL) on 14 days during that had the maximum Austin concentration at Dripping Springs Figure Inter-state back-trajectories (based on 5-day HYSPLIT back-trajectories initiated at 1 km AGL) on 22 days during that had the maximum Austin concentration at Round Rock, Lake Georgetown, or Hutto Figure 9-6. Inter-state back-trajectories (based on 5-day HYSPLIT back-trajectories initiated at 1 km AGL) on 16 days during that had the maximum Austin concentration at San Marcos or McKinney Roughs Page 16 of 211

17 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area 1 Introduction As part of its ongoing air quality planning efforts, CAPCOG regularly updates the ozone conceptual model for the Austin-Round Rock Metropolitan Statistical Area (MSA), which consists of Bastrop, Caldwell, Hays, Travis, and Williamson Counties. The region s ozone conceptual model characterizes the area s ozone problem both qualitatively and quantitatively, and is one of the primary technical analyses used for regional ozone planning. This report uses EPA s guidance for developing ozone conceptual models, which can be found in its 2007 modeling guidance 4 and is 2014 draft modeling guidance. 5 The analyses are similar to the analyses included in CAPCOG s previous ozone conceptual models for the region, which were developed in in 2004, , , 8 and These conceptual models analyzed air pollution and meteorological data covering the following periods: 2004 Conceptual Model: ; 2007 Conceptual Model: ; 2010 Conceptual Model: ; and 2012 Conceptual Model: Each of these conceptual models analyzed 8-hour ozone concentrations that exceeded certain thresholds related to the EPA s National Ambient Air Quality Standards (NAAQS) for ozone. The 2004, 2007, and 2012 conceptual models were focused on days when 8-hour ozone concentrations were at or above 75 parts per billion (ppb), while the 2010 conceptual model analyzed days that were at or above three levels of a potentially more stringent ozone standard that EPA was considering at the time: 70 ppb, 65 ppb, and 60 ppb. This conceptual model covers ozone monitoring data, and analyzes the conditions that lead to high 8-hour ozone concentrations within the region. High Ozone is defined in this conceptual model in five different ways: 1. Days when peak 8-hour ozone averages exceed 75 ppb; 2. Days when peak 8-hour ozone averages exceed 70 ppb; 3. Days when peak 8-hour ozone averages exceed 65 ppb; 4. The four highest daily peak 8-hour ozone averages each year; 5. The ten highest daily peak 8-hour ozone averages each year. 75 ppb was selected as a threshold for analysis since the 2008 ozone NAAQS that is currently in effect is set at 75 ppb. 70 ppb and 65 ppb represent the high and low ends of the range of EPA s 2014 proposal to revise the ozone NAAQS. For this analysis, a day was considered a high ozone day for the region if at least one ozone monitor measured a peak eight-hour ozone concentration above these thresholds AUS/04086sipapg_pro.pdf Austin_Area_Conceptual_Model_2012.pdf Page 17 of 211

18 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area The use of a numeric threshold has some limitations in that the meteorological conditions that might lead to an exceedance of a 75 ppb threshold several years ago may only lead to an exceedance of a 70 ppb threshold today after several years of emission reductions. Therefore, the analysis of the four-highest and ten-highest daily peak 8-hour ozone averages at the region s two regulatory ozone monitors, which have operated year-round over this entire period, provides a different perspective on the conditions that occur on the region s worst ozone days, regardless of the absolute level of the 8-hour averages. The four highest levels each year are important because they are used for assessing compliance with the ozone NAAQS, and the ten highest levels are relevant for modeling, since EPA s most recent draft ozone modeling guidance recommends the use of the ten highest modeled 8-hour ozone concentrations for attainment modeling rather than using only days above a numeric threshold. 10 Whereas CAPCOG contracted with the University of Texas at Austin (UT) to complete the previous ozone conceptual models, this report is the result of collaboration between CAPCOG and the Alamo Area Council of Governments (AACOG). CAPCOG completed the analyses in sections 3, 4, and 5, while AACOG completed the analyses in sections 6, 7, 8, and 9. Each party quality-assured the other s work consistent with the projects level 3 quality assurance project plan (QAPP). A QA report is included as an appendix to this report. The conceptual model presented in this report is specific to ozone concentrations averaged over 8 hours. For brevity, ozone concentration(s) is sometimes used in place of ozone concentration(s) averaged over 8 hours throughout the text discussions presented in this document. If a different averaging period is relevant (i.e. 1-hour), it will be mentioned specifically. At times throughout this report, the terms Austin area, Austin-Round Rock area, Austin-Round Rock MSA, and Central Texas are used to describe the region analyzed in this conceptual model. All of these terms should be taken to mean the five-county Austin-Round Rock MSA, as defined by the Office of Management and Budget (OMB) in 2013, which includes Bastrop, Caldwell, Hays, Travis, and Williamson Counties. 1.1 Conceptual Model Definition According to the EPA s Ozone Modeling Guidance, a conceptual model of ozone formation is a qualitative way to characterize the nature of an area s ozone problem through a comprehensive look at the influence of emissions, meteorology, transport, and other relevant atmospheric processes on air quality in the area. The conceptual model should, at minimum, summarize both the local meteorological conditions and associated large-scale weather patterns typically experienced during periods of elevated ozone. In addition, the supporting analysis should include a review of available ambient air quality data, meteorological data, and previous photochemical modeling efforts. A separate project CAPCOG has completed includes a comprehensive analysis of previous photochemical modeling efforts, so this analysis is not included in this report in order to avoid duplication. This report includes a description of the necessary conditions for high ozone concentrations in the Austin area. Based on previous analyses for Austin, a set of sufficient conditions for the formation of high ozone (i.e., conditions under which high ozone will always occur) has not been technically feasible. CAPCOG will also address other relevant questions listed in Section of EPA s ozone modeling guidance, Guidance on the Use of Models and Other Analyses to Demonstrating 10 Page 18 of 211

19 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area Attainment of Air Quality goals for Ozone, PM2.5, and Regional Haze 11 and Section 2.1 of EPA s draft December 2014 update to the guidance 12. This guidance document suggests several key elements to include in a model: Describe the ambient monitoring network used for the conceptual model: This is addressed in Section 2. Describe the status and trends of air quality in the area: This is addressed in Section 3 Investigate possible relationships between emissions and air quality: This is addressed in Section 1.3, although not all topics are covered in this conceptual model. Investigate possible relationships between meteorology and air quality: This is addressed in Section Austin-Round Rock MSA Ozone Levels The Clean Air Act (CAA) of 1990 requires the U.S. Environmental Protection Agency (EPA) to establish NAAQS for pollutants considered harmful to public health and the environment. Primary standards are set to protect public health, including "sensitive" populations; secondary standards are set to protect public welfare. The current primary NAAQS for ozone is ppm (75ppb) averaged over 8 hours. The CAA requires NAAQS review on a five-year cycle, and in 2014, EPA proposed a more stringent 8-hour ozone NAAQS be set at ppb. EPA is expected to finalize the new standard by October 1, This conceptual model examines both ends of this proposed range, as well as the current standard level. The Texas Commission on Environmental Quality (TCEQ) operates two ozone monitoring stations in Travis County to fulfill federal requirements for ozone monitoring in a metropolitan area the size of the Austin-Round Rock MSA. EPA uses the ozone monitoring data at these stations continuous air monitoring station (CAMS) 3 and CAMS 38 to determine whether the Austin-Round Rock MSA, is in compliance with the ozone NAAQS. An area is considered to be violating the primary ozone NAAQS if the annual fourth highest 8-hour daily maximum concentration, averaged over three consecutive years, exceeds the standard at any regulatory monitoring station within the region. This statistic is known as the region s ozone design value. The Austin-Round Rock MSA s most recent design value for was 69 ppb, which is 92% of the level of the 2008 ozone NAAQS. Figure 1-1 shows the annual 8-hour design values for Austin for years As the figure shows, the region s 8-hour ozone design values have steadily decreased over these years, and the region has managed to narrowly avoid being designated nonattainment for ozone in 2004 and However, while the Austin area s 2014 ozone design value would have met a 70 ppb standard, it would not meet a 65 ppb standard if EPA sets it that low, and will likely be above 65 ppb for the period EPA has indicated it plans to use as the basis for nonattainment designations for the new standard. As Figure 1-1 below shows, the Austin-Round Rock MSA s ozone design value declined sharply in This was based on the very low ozone levels measured in 2014, with a fourth highest daily maximum 8-hour ozone average of only 63 ppb at CAMS 38. From 1999 to 2014, the design value decreased by about 1.3 ppb each year, while the decrease between 2013 and 2014 was more than three times that Page 19 of 211

20 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area Figure 1-1. Ozone Design Value Trend for the Austin-Round Rock MSA, y = -1.31x R² = Hour Ozone Design Value (ppb) Relationships Between Emissions and Air Quality This subsection addresses the topics that EPA s 2014 draft modeling guidance indicate should be covered in a conceptual model pertaining to the relationship between emissions and air quality. As was described earlier, about ppb of peak 8-hour ozone concentrations can be attributed to natural conditions and international ozone transport. Of the remaining portion of the 8-hour ozone levels attributable to anthropogenic emissions from the United States, about a third can be attributed to local emissions within the MSA, a third to the rest of Texas, and a third to the rest of the U.S., based on modeling conducted by the University of Texas at Austin in Sensitivity modeling of broad reductions in local NO X and VOC emissions has shown that the area s ozone levels are times more sensitive to anthropogenic NOX emissions than VOC emissions Key Emission Sources or Source Categories Within the MSA, the following table shows the total emissions reported in the 14 main tiers of sources in the 2011 National Emissions Inventory. About 60% of the region s NO X emissions come from on-road sources, 20% from non-road sources, 5% from electric generating units, and the remaining 15% from other stations sources. About 40% of the region s VOC emissions come from miscellaneous, which includes consumer products, 20% comes from on-road sources, 20% from solvent utilization, 10% from oil and gas, and the remaining from other sources. The very high quantity of emissions from miscellaneous sources in 2011 is mainly attributable to the severe wildfires that occurred in the region in September of that year these emissions are very atypical and were indeed historically bad, with the wildfire in Bastrop County counted as one of the worst in the state s history Precursor_Response_Runs_Final.pdf Page 20 of 211

21 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area Table Annual Emissions by Tier 1 Source Category (tons per year) TIER1 TIER1NAME NO X VOC 1 FUEL COMB. ELEC. UTIL. 2, FUEL COMB. INDUSTRIAL 1, FUEL COMB. OTHER 1, PETROLEUM & RELATED INDUSTRIES 937 6,373 6 OTHER INDUSTRIAL PROCESSES 2, SOLVENT UTILIZATION 0 13,462 8 STORAGE & TRANSPORT 5 2,355 9 WASTE DISPOSAL & RECYCLING HIGHWAY VEHICLES 26,046 13, OFF-HIGHWAY 8,882 4, MISCELLANEOUS 2,003 29, CHEMICAL & ALLIED PRODUCT MFG 0 3 All TOTAL 45,768 71, Comparison of Emission Trends for Annual and/or Seasonal/Episodic Periods to Corresponding Air Quality Trends A comparison of photochemical modeling emissions inventories for the region show that there has been a decrease in weekday NO X emissions of about 27% from within the MSA from 2006 to The following table shows the average weekday (Monday-Thursday) emissions in the June 2006 base case photochemical modeling episode, 15 along with the average weekday emissions in the 2012 baseline scenario modeled by AACOG in Table 1-2. Comparison of 2006 and 2012 photochemical modeling NOX emissions inventories for the MSA County 2006 (tpd) 2012 (tpd) Difference (tpd) Difference (%) Bastrop % Caldwell % Hays % Travis % Williamson % TOTAL % This reduction in NO X emissions would be expected to account for a 3.1 ppb reduction in peak 8- hour ozone concentrations at the region s two regulatory ozone monitors based on UT s sensitivity modeling. The actual reduction in the design values at CAMS 3 and CAMS 38 over this time frame was 8 ppb at both sites (a 10% reduction), meaning that local NO X reductions 15 Base_case_Performance_Evaluation_Final.pdf 16 Body_Only.pdf Page 21 of 211

22 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area accounted for 38-39% of the measured reduction in the region s ozone design value over that time period. This decrease in the region s design value was greater than the average decrease in design values for all monitoring stations in Texas (7 ppb, or 8%) and the average decrease in design values for all monitoring stations across the country (4 ppb, or 5%) over this same time period. 17 Within these years, a comparison of the 2008 and 2011 NEI data provides an opportunity to compare the extent to which emission reductions from within the MSA were occurring more or less quickly than emissions at the state or national level. There was a 10% reduction in NO X emissions between 2008 and 2011 in the MSA, compared to 16% within the state of Texas as a whole and 15% across the U.S. While, for some important source categories, the methods used to estimate the region s emissions changed over this time (such as for locomotives), which accounts for some of the difference, it may be true that the MSA s emissions are not declining as rapidly as the state or the nation as a whole. One reason for this might be the region s tremendous growth that has occurred in recent years. Whereas Texas s population grew by 5.5% between 2008 and 2011, the Austin-Round Rock MSA s population grew by 8.0%. 18 Table 1-3. Comparison of MSA, Texas, and U.S. NO X emissions in 2008 and 2011 (minus miscellaneous ) Area 2008 NO X Emissions (tpy) 2011 NO X emissions (tpy) Difference (tpy) % Difference Austin-Round Rock MSA 48,382 43,764-4,617-10% Texas 1,515,260 1,266, ,820-16% USA 16,647,844 14,119,439-2,528,405-15% Over this same time period, the average design value for all ozone monitors across the country decreased by 5 ppb (7%), while the average design value for monitors in Texas decreased by 2 ppb (3%) and the average design value for Travis County decreased by 3 ppb (4%) Resources on Other Topics on Relationship between Emissions and Air Quality In addition to the topics covered above, the 2014 draft modeling guidance recommends the following analyses: Assess Historical Effectiveness of Control Programs Consider How Future Emissions Growth or Reductions May Affect Air Quality List Control Programs That Are in Place or That Will Soon be Implemented That May Impact Emissions Sources in the Area CAPCOG is not including a discussion of these topics in this conceptual model because these topics are more extensively covered in other recent documents CAPCOG has produced, including the region s annual air quality report and an analysis of regional photochemical modeling data Page 22 of 211

23 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area 1.4 Local Air Quality Planning The Austin-Round Rock MSA is one of ten areas in Texas that receives funding from the Texas Legislature to address ozone air quality issues through the near nonattainment area (NNA) grant program. The Capital Area Council of Governments (CAPCOG) coordinates air quality planning activities in the five-county Austin-Round Rock MSA. As part of these efforts, the Austin-Round Rock MSA has participated in four different voluntary air quality planning efforts designed to keep the area in attainment of the ozone NAAQS: The One-Hour Ozone Flex Program (2002); The Early Action Compact ( ); The Eight-Hour Ozone Flex Program ( ); and The Ozone Advance Program (OAP) (2012-present). The most recent air quality plan adopted for the region by the Central Texas Clean Air Coalition (CAC) was the OAP Action Plan, which was adopted in December The OAP Action Plan is designed to achieve five goals: 1. Stay in attainment of the 2008 eight-hour ozone NAAQS of 75 ppb; 2. Continue reducing the region s 8-hour ozone design value to avoid being designated nonattainment for a new ozone NAAQS; 3. Put the region in the best possible position to bring the area into attainment of an ozone standard expeditiously if it is does violate an ozone standard or gets designated nonattainment; 4. Reduce the exposure of vulnerable populations to air pollution when the region experiences high ozone levels; and 5. Minimize the costs to the region of any potential future nonattainment designation. This conceptual model enables the CAC and the region s air quality planners to better understand the dynamics of Central Texas air quality and therefore to better address these goals. 1.5 Relevance of Conceptual Model for Future Photochemical Modeling The development of air quality plans involves a number of complex steps that include ambient air quality data analyses, emissions inventory development, air quality modeling, and analysis of emissions control options. An ozone conceptual model is an important part of this process, and is essential for photochemical modeling. While typically, photochemical modeling is only done for areas that are designated nonattainment for an ozone NAAQS, CAPCOG conducts photochemical modeling for the Austin-Round Rock MSA as part of the region s efforts to stay in attainment of the NAAQS. A crucial part of the process in photochemical grid modeling is selection of one or more multi-day high ozone episodes or episodes that cover one month or an entire ozone season. It is essential that the episodes selected for photochemical grid modeling be representative of weather patterns that are most often associated with high ozone concentrations measured in the local area. As such, the development of a conceptual model of the large-scale weather patterns and associated local meteorological conditions that are most often observed during high ozone episodes is the first step in episode selection. Page 23 of 211

24 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area Historically, a limited number of days (typically from three to ten) were selected for photochemical modeling used to demonstrate that future ozone concentrations would be in compliance with the 8-hour ozone NAAQS. With the advancement of computer technology over the past decade, computer speed and storage issues are less of an impediment for modeling relatively long time periods. In EPA s 2007 modeling guidance, EPA suggests that episodes for photochemical modeling cover, at minimum, a full synoptic cycle, which may be anywhere from 5-15 days. Ideally, modeling even longer time periods of up to a full season may simplify the episode selection process and provide a rich database with which to apply the modeled attainment test. 19 The latest draft EPA guidance for episode selection favors the simulation of a relatively long period (i.e., months to a year or longer) that includes a number of high ozone days. The previous version of this conceptual model 20 discussed the June 2006 base case. Section 9 of this model discusses the meteorological conditions for fall 2006, fall 2011, and June 2012, other periods of high ozone in the region that have been modeled. 2 Data Used in This Conceptual Model This ozone conceptual model for the Austin-Round Rock MSA is based on ozone and meteorological data for This section describes the datasets that were used to develop the conceptual model. These datasets include ozone and meteorological data collected at monitoring stations operated by TCEQ, the National Weather Service (NWS), and CAPCOG. CAPCOG investigated local and regional ozone transport using back-trajectories. Regional transport was investigated using the NOAA back-trajectory models provided by the Earth System Research Laboratory (ESRL) and HYSPLIT (Hybrid Single-Particle Lagrangian Integrated Trajectory) model. 2.1 Regional Ozone Monitoring Data A total of 14 CAMS collected ozone data between 2006 and Two of these stations are Federal Reference Method (FRM), regulatory monitors operated by TCEQ (CAMS 3 and 38). These two monitoring stations operated year-round for the entire period covered by this study. The other 12 CAMS were operated by CAPCOG for varying durations, some covering the entire period (CAMS 614) with others covering less than one full ozone season (Liberty Hill and Elroy) Austin_Area_Conceptual_Model_2012.pdf Page 24 of 211

25 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area Figure 2-1. Map of ozone CAMS in the Austin-Round Rock MSA active between From 2006 to 2014, the following changes to the ozone monitoring network occurred: CAMS 674 (Round Rock) started operations in June 2006 and ended at the end of the 2010 ozone season; CAMS 675 (San Marcos) started operations in June 2006 and ended in September 2011; CAMS 613 (Pflugerville) ended operations at the end of the 2006 ozone season; CAMS 690 (Lake Georgetown) started operations in May 2008; Page 25 of 211

26 Conceptual Model for Ozone in the Austin-Round Rock Metropolitan Statistical Area CAMS 6602 (Hutto) started operations in May 2011; CAMS 1675 (San Marcos Staples Road) started operations in September 2011; Liberty Hill temporary CAMS operated from May October 2012 (did not report to LEADS); Elroy temporary CAMS operated from June October 2012 (did not report to LEADS); CAMS 1604 (Lockhart) began operation in May 2013 and began reporting to LEADS at the beginning of the 2014 ozone season; CAMS 1603 (Gorzycki Middle School) began operation in September 2013 and began reporting to LEADS at the beginning of the 2014 ozone season; and CAMS 6602 (Hutto) sampling equipment was moved to a different location at the same campus. CAMS 1675, operational since September 2011, is approximately 2 miles south-southeast of the original location. Given the close distance between CAMS 675 and CAMS 1675, current San Marcos monitoring locations and the short measurement history at CAMS 1675, most analyses in this report (e.g., ozone trends and back-trajectories), combine the observations from CAMS 675 and CAMS 1675 into a single dataset. The Hutto monitoring station has also changed locations, albeit at the same address. In 2011 it was originally placed in an unused room, but due to the facility owner s need to expand, it was relocated to a different building on campus at the beginning of the 2014 ozone season and mounted on the roof. Throughout the 2014 ozone season, this ozone analyzer measured ozone levels far below all of the other ozone concentrations measured throughout the region and below the levels that had previously been measured at CAMS 6602 from While, after conducting co-located sampling, CAPCOG determined that the instrument s measurements were accurate, CAPCOG and CAPCOG s monitoring contractor (Dios Dado Environmental, Ltd.) determined that the abnormally low ozone measurements could only be attributed to the siting of the ozone monitor, and were not likely representative of the ground-level concentrations in the community. Therefore, CAPCOG excluded Hutto 2014 ozone monitoring data from this analysis. Table 2-1 provides a summary of basic information for each of these ozone monitoring stations. As of 2015, there are nine active ozone monitoring stations within the Austin-Round Rock MSA: CAMS 3, 38, 614, 684, 690, 1603, 1604, 1675, and CAPCOG obtained ozone and meteorological data from TCEQ s Texas Air Monitoring Information System (TAMIS) and from TCEQ s Leading Environmental Analysis and Display System (LEADS). For the Liberty Hill and Elroy sites, and for the Gorzcyki and Lockhart sites in 2013, CAPCOG used data downloaded directly from the stations data loggers, since these data were not put into LEADS or TAMIS. Page 26 of 211

27 Table 2-1. Description and location of ozone monitoring stations in the Austin-Round Rock MSA Monitoring Station Name CAMS # AIRS ID County Latitude Longitude Agency Operational Period (since 2006) Austin Northwest/ Murchison Travis TCEQ Present Audubon Travis TCEQ Present Dripping Springs* Hays CAPCOG March Present McKinney Roughs* Bastrop CAPCOG Aug 16, Present Pflugerville* Travis CAPCOG March 2006 Oct 2006 Round Rock* Williamson CAPCOG June 2, 2006 Oct 2010 Lake Georgetown* Williamson CAPCOG May 17, Present Liberty Hill* N/A N/A Williamson CAPCOG May 30, 2012 Oct 2012 Elroy* N/A N/A Travis CAPCOG June 28, 2012 Oct 2012 Hutto* Williamson CAPCOG May 18, 2011 Present San Marcos* Hays CAPCOG June 15, 2006 Sept 14, 2011 San Marcos Staples Rd.* Hays CAPCOG Sept 20, Present Lockhart* Caldwell CAPCOG May 29, Present Gorzycki* Travis CAPCOG Sept 25, Present *Ozone seasonal monitoring stations. The ozone transport analyses presented in Section 5 of this report are based on Austin area back-trajectories calculated from meteorological observations of surface wind speed and wind direction at these ozone monitoring stations. The back-trajectories represent the hypothetical path of a near-surface air parcel assuming the parcel moves with the area-averaged winds as measured by the Austin area network of surface monitoring stations. Page 27 of 211

28 2.2 National Weather Service Synoptic Analyses The conditions conducive to the transport, formation, and accumulation of ozone in Central Texas are primarily dependent on the prevailing large-scale weather patterns. The large-scale atmospheric circulation features during high ozone episodes in Austin for were primarily investigated using upper air and surface weather maps maintained by the Precipitation Diagnostics Group in the Mesoscale and Microscale Meteorology Division of National Center of Atmospheric Research (NCAR) 21. The NWS provides both surface weather maps and upper-air charts for the continental U.S. on a near real-time basis. These maps summarize data analyses that incorporate meteorological observations at airports and automated observing platforms, radiosondes, radar scans, and present weather conditions from monitoring sites across the U.S. Contours of temperature, dew-point temperature, pressure, wind direction, and wind speed were used to investigate the locations of large-scale synoptic features such as frontal troughs, high pressure ridges, and the jet stream. All weather charts analyzed or discussed in this report were obtained from the Precipitation Diagnostics Group in the Mesoscale and Microscale Meteorology Division of NCAR or the UNISYS Weather website 22. Twice-daily visible, infrared, and water vapor imagery obtained from the NOAA GOES satellite were also reviewed via the UNISYS website. 2.3 HYSPLIT Back-Trajectories The Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model was used to investigate the potential source regions of air entering the Austin area. HYSPLIT was developed by a joint effort between the National Oceanic and Atmospheric Administration (NOAA) and Australia s Bureau of Meteorology, and can be run interactively on the World Wide Web 23. HYSPLIT uses meteorological model forecast data from the National Centers for Environmental Prediction (NCEP) archived by Air Resources Laboratory (ARL). All trajectories were based on the three-dimensional wind field provided by the Eta Data Assimilation System (EDAS). The EDAS data are archived by ARL with a 3-hour time interval over a 40-km grid. Additional information on the meteorological data archive can be found on the NOAA website 24. It is important to note that the surface and HYSPLIT back-trajectories shown in this report may have substantial uncertainty. Surface back-trajectories are based on a relatively small dataset that may not capture important spatial and temporal variability in winds at locations away from the available monitoring stations. In addition, the surface back-trajectories assume that a given parcel moves with surface winds as measured by the network of monitoring stations throughout the back-trajectory period, and does not account for any vertical movement of air parcels. The vertical movement of air can have a substantial impact on the transport of emissions from a stationary source, for example, particularly if vertical motions move the air parcel into a layer of the atmosphere characterized by a different wind speed and wind direction compared to that measured near the surface. For the HYSPLIT back-trajectories, the uncertainty primarily arises from the three-dimensionally gridded meteorological dataset used in the back-trajectory calculation. The meteorological data are obtained by merging available weather forecast and observational data. Sources of uncertainty include, but are not limited to, errors associated with the available meteorological observational data, performance of the Page 28 of 211

29 meteorological forecast models, and errors introduced by the data assimilation technique used to construct the merged dataset. In addition, the relatively low horizontal spatial resolution and temporal resolution probably captures the prevailing synoptic flow during weather patterns dominated by largescale and slow-moving systems that are well-simulated by the meteorological models. The EDAS archive may not properly capture the impacts of localized or mesoscale weather features, such as the land-sea breeze along coastal Texas, whether or not these features are well-simulated by the meteorological models. Uncertainty is introduced by errors in the simulation of large-scale vertical motion by the meteorological models in combination with the vertical resolution of the archived datasets (approximately 250 meters in the lower atmosphere for the EDAS datasets). In addition, the calculation of the back-trajectories can be sensitive to the starting height above ground depending on the change of wind direction and wind speed with respect to height above the surface over the geographic areas of interest. Further discussion of the sources of uncertainty associated with the calculation of trajectories using the HYSPLIT model can be found at the HYSPLIT website 25. In general, the user is advised that the HYSPLIT back-trajectories presented in this report should not be interpreted as precise tracks of air parcels; however, patterns that emerge when analyzing a relatively large number of back-trajectories should provide a good indication of potential long-range transport due to the prevailing large-scale flow regime. 3 Temporal and Spatial Patterns in High Ozone Levels in the Region This section includes an analysis of the temporal and spatial variability of ozone within the region, including trends in ozone levels over time. Throughout the analyses presented in this report, data interpretations are inherently biased by the locations and number of the individual sites within the monitoring network. As discussed below, the number and locations of the ozone monitors in the Austin-Round Rock area changed considerably from 2006 through 2014 and the existing monitors at a given time may not always be ideally situated to sample the upwind or downwind urban plume. Therefore, the maximum ozone concentrations in the area may have been substantially different than the available data indicate. Everything else being equal, a monitoring network with more stations would measure more high-ozone days than the same network with fewer stations. Therefore, the monitoring network in 2014, which included nine ozone stations, is very likely more capable of measuring high ozone concentrations on any given day than the monitoring network could at the beginning of 2006, when only four ozone monitoring stations were operating (Audubon, Murchison, Dripping Springs, and Pflugerville). In mid-june 2006, CAPCOG added ozone seasonal monitoring at Round Rock and San Marcos, followed by McKinney Roughs in late summer. McKinney Roughs, located approximately 20 miles southeast of downtown Austin, is well-positioned to measure background ozone concentrations entering the Austin area during periods with southeasterly near-surface winds. After the 2006 ozone season, the Round Rock monitor essentially replaced the Pflugerville monitor and measured background ozone concentrations entering Austin when winds are from the north or northeast, and ozone concentrations downwind of the Austin urban area during southerly winds. In 2008, CAPCOG installed an additional ozone season monitor on the north side of Austin at Lake Georgetown and the Round Rock monitor was re-located to Hutto in In September 2011, CAPCOG moved the San Marcos monitoring station to a new location on the eastern side of the city, which is approximately 2 miles south-southeast of the 25 Page 29 of 211

30 original location. Given the close distance between the original and current San Marcos monitoring locations and the short measurement history at CAMS 1675, most analyses in this report (e.g., ozone trends and back-trajectories), combine the observations from CAMS 0675 and CAMS 1675 into a single dataset. In 2012, two monitoring sites were tested at Liberty Hill, northwest of CAMS 38 (Audubon) and east of Georgetown, and Elroy, slightly west of McKinney Roughs. They were discontinued after the 2012 ozone season once data collection had concluded. In 2013, CAPCOG set up two temporary stations in Lockhart, in Caldwell County, and at Gorzycki Middle School in southwest Travis County. In 2014, these two stations were made into permanent stations and began reporting to LEADS. 3.1 Location of Daily Peak Ozone Concentrations During , one or more Austin monitors measured maximum ozone concentrations greater than 65 ppb on 254 unique dates. Figure 3-1 presents the number of days that each area monitor measured the maximum area-wide ozone concentration for different ozone levels. The monitors are ordered from left to right from the longest running to the shortest running sites. It is important to note that not all of these monitors were operational Also note the areawide daily maximum may have been at more than one location and two monitors may have measured the same concentration. While Audubon and Murchison sites had the most number of area-wide high ozone concentrations, to some extent, this reflects the fact that these monitors have been online for the full nine years and run year-round while the others operate seasonally and most were not operational for this entire period. The data for the two regulatory monitors represented the region-wide peak 8- hour ozone averages on 48% of the days when at least one station measured over 65 ppb, 51% of the days when at least one station measured over 70 ppb, and 48% of the days when at least one monitor measured over 75 ppb. Figure 3-1. Number of area-wide maximum ozone concentrations measured at each ozone monitoring station >65 ppb >70 ppb >75 ppb Page 30 of 211

31 Since not all monitors were operations for this entire period, it is useful to normalize this analysis based on the number of ozone seasons that the monitor was operational. The following figure normalizes these data by dividing the number of days the monitor measured the region-wide maximum 8-hour ozone concentration by the number of ozone seasons that the monitor was operational to obtain the average number of region-wide maximum measurements each monitor would measure in an average ozone season during this period. Finally, dividing that number by the overall average number of high ozone days that occurred each year over this period provides a normalized likelihood of a monitoring station measuring the highest 8-hour ozone average for a high ozone day. The data for Lockhart, Gorzycki, Pflugerviile, Liberty Hill, and Elroy in this analysis should be taken with a grain of salt due to the low number of ozone seasons these stations were operational over this period. Figure 3-2. Likelihood that a station could measure the highest 8-hour ozone average on a high ozone day, >65 ppb >70 ppb >75 ppb 35% 30% 25% 20% 15% 10% 5% 0% 3.2 Fourth Highest Eight-Hour Ozone Concentrations Since not all days above a numeric threshold are necessarily used for comparing a region s ozone levels to the NAAQS, it is also useful to analyze the spatial variability. The table and figure below shows the spatial variability of these data for each monitor that was operational in 2014 in ozone levels that are directly comparable to the form of the NAAQS, the three-year average of the fourth-highest daily maximum eight-hour ozone concentrations. Page 31 of 211

32 Table th Highest daily maximum 8-hour ozone concentrations (ppb) for Austin-Round Rock MSA Monitors, Monitor th -Highest 8-Hour Ozone th -Highest 8-Hour Ozone th -Highest 8-Hour Ozone Average Audubon (C3) Murchison (C38) Dripping Springs (C614) San Marcos (C1675) McKinney Roughs (C684) Georgetown (C690) Gorzycki (C1603) N/A N/A 57 N/A Lockhart (C1604) N/A N/A Hutto (C6602) * 59* *Hutto monitoring station produced consistently low ozone measurements throughout the 2014 ozone season, and was relocated on the property in 2015 as a result. The 3-year average for Hutto is therefore skewed low. Page 32 of 211

33 Figure th highest 8-hour ozone concentrations at Austin-Round Rock MSA CAMS Page 33 of 211

34 3.3 Frequency of High Ozone Days by Year The trend in annual high ozone days must be interpreted with caution since the locations and numbers of Austin area monitors have changed throughout the period. For example, the Austin monitoring network was expanded from seven stations at the end of 2006 (Audubon, Murchison, Dripping Springs, Pflugerville, Round Rock, San Marcos, McKinney Roughs), to nine stations by the end of the 2014 ozone season with the inclusion of monitors at Gorzycki Middle School in Austin, Lake Georgetown, Lockhart, and Hutto. Within this time period, the number of monitors operational varied from two at the edge of the seasons (Audubon & Murchison) to different combinations of nine. As shown in the figure below, the number of high ozone days over 65 ppb (based on all Austin area monitors) varies from 50 in 2006 to 6 days in The amount of high-ozone days varies greatly yearto-year, with an overall downward trend. The spike in the number of high ozone days in 2011 coincided with exceptional meteorology and numerous large-scale wildfires that increased ozone levels across the state. Generally, the overall trend has been downward, especially at higher ozone levels. Considering the fact that there were more than twice the number of monitors from than there were at the beginning of 2006, these trends likely understate the actual decline for the region over this period. Figure 3-4. Annual number of high ozone days at one or more Austin area monitor, Number of Days > 65 ppb > 70 ppb > 75 ppb There are also some substantial differences in the number of high ozone days based on measurements from regulatory monitors only compared to all available Austin area monitoring data. For example, in 2011 the regional monitors cumulatively measured 20 days with ozone concentrations over 70ppb, the two regulatory monitors only measured 5 such days was highly unusual in that area-wide maximum ozone concentrations were measured on three days at Dripping Springs and six days at San Marcos. During , San Marcos never measured the Austin area maximum ozone concentration, while the maximum was measured at Dripping Springs on one day during 2009 and two days in Figure 3-5, 3-6, and 3-7 display the annual trends comparing the number of days in which the regulatory monitors vs. CAPCOG monitors measured the peak maximum ozone concentration in the area. To some extent, the larger number of days when CAPCOG s ozone monitors measure the peak-hour ozone average for the region on high ozone days in recent years reflects the growth in CAPCOG s monitoring network from two stations in the MSA in 2006 to seven in However, since the TCEQ monitors are Page 34 of 211

35 sited to measure the region-wide peak ozone concentrations downwind from the urban core, this trend may also reflect the decreased importance of urban emissions and the increased importance of background ozone levels Austin to the region s peak 8-hour ozone averages. Figure 3-5. Maximum 8-hour ozone concentration >75ppb measured at regulatory and CAPCOG monitors, Number of Days CAPCOG Monitors Only Regulatory Monitors Only Figure 3-6. Maximum 8-hour ozone concentration >70ppb measured at regulatory and CAPCOG monitors, Number of Days CAPCOG Monitors Only Regulatory Monitors Only Page 35 of 211

36 Figure 3-7. Maximum 8-hour ozone concentration >65ppb measured at regulatory and CAPCOG monitors, Number of Days CAPCOG Monitors Only Regulatory Monitors Only Frequency of High Ozone Days by Month The monthly distribution of high ozone days during is shown in Figure 3-8. The monthly frequency of occurrence of high ozone days is characterized by a bi-modal distribution, with a relative minimum in July and highest numbers of ozone days during May-June and August-October. Measurements over 65 ppb were recorded as early as March and as late as November, while measurements over 75 ppb were only recorded between May and October. Figure 3-8. Monthly distribution of high ozone days for all monitoring sites, % 30% 25% 20% 15% 10% 5% >65 ppb >70 ppb >75 ppb 0% Given the bi-modal distribution of high ozone days shown above, CAPCOG divided the ozone season into early (March June) and late (August November) season days, eliminating July to have an even number of months. Figure 3-9 shows the number of days with ozone concentrations greater than 70 ppb Page 36 of 211

37 in each period as an example. While there is a clear seasonal difference between the two periods, there is no evident pattern that more or fewer high ozone days occur in the early or late season relatively. In 2006, 2008, 2009 and 2012, there were more days with ozone levels over 70 ppb before July than after it, and vice versa for 2007, 2010, 2011, and There were no days over 70ppb in Figure 3-9. Annual number of early and late season days > 70 ppb, Number of Days >70ppb Early Season (Mar-June) Late Season (August- November) When examining just the highest ozone concentrations per year at each regulatory monitor, the distribution between early and late season high ozone days is fairly even, as exhibited in the figure below that includes both the highest 4 and highest 10 ozone concentrations per monitor. There is still typically a peak in May and September, with a relatively lower number of high ozone days in July. Figure Monthly distribution of highest four ozone days for CAMS 3 & CAMS 38, % 30% 25% 20% 15% 10% CAMS 3 CAMS 38 5% 0% Page 37 of 211

38 Figure Monthly distribution of highest ten ozone days for CAMS 3 & CAMS 38, % 20% 15% 10% 5% CAMS 3 CAMS 38 0% These analyses suggest that, while in any given year high ozone days may be clustered more in the earlier or later part of the year, there does not appear to be clear evidence that the earlier part of the season is typically worse than the later part or vice-versa over several years, and across each of the five types of days analyzed. 3.5 Frequency of High Ozone Days by Day of the Week In general, the number of high ozone days by day of the week show large variability among the monitoring sites and suggests that day-to-day differences are associated with normal variability and a result of many other factors. Because meteorological factors such as high temperatures or specific weather patterns are not correlated to the day of the week, but are better correlated to the month of the year, one would expect to see much less correlation between day of the week to ozone levels. However, there are some patterns worth noting in the data from the last nine years. The numbers of high ozone day occurrences by day of the week is presented in Figure There is an upward trend from Sunday toward the end of the week and into the weekend. Interestingly, while Sunday is clearly the day with the lowest number of high-ozone days for all thresholds, ozone levels above 65 ppb or 70 ppb most frequently occur on Saturdays, while ozone levels above 75 ppb most frequently occur on Fridays. Predictably, there are more days over 65 ppb than over 75 ppb, but their overall rise is similar. Page 38 of 211

39 Figure Number of high ozone days by day of the week at all Austin area monitors, Number of Days Sunday Monday Tuesday Wednesday Thursday Friday Saturday 0 >65 ppb >70 ppb >75 ppb Aside from Sunday being the day least likely to record high ozone, it is difficult to draw any clear conclusions from data for Monday Saturday when looking at the number of high ozone measurements by day of the week at the regulatory monitors only (CAMS 3 and 38). In the figure below it is apparent that there is more variability in day of the week for days when ozone levels were greater than 75ppb than for the lower thresholds of >70 ppb and >65 ppb. Figure Number of high ozone measurements by day of week at regulatory monitors CAMS 3 & 38 only, Number of Days Sunday Monday Tuesday Wednesday Thursday Friday Saturday 0 >65 ppb >70 ppb >75 ppb It is also possible to examine CAMS 3 and CAMS 38 s four and ten-highest ozone days per year, shown in Figure 3-14 and Figure The two analyses display similar patterns, but again with less variability day Page 39 of 211

40 to day. The high percentage of the four highest 8-hour ozone concentrations that occur on Fridays and low percentage that occur on Sundays is consistent with the tendency of on-road emissions to be highest on Fridays and lowest on Sundays. Figure Day of the week distribution of highest four ozone days for CAMS 3 & CAMS 38, % 20% Frequency of Days 15% 10% 5% CAMS 3 CAMS 38 0% Figure Day of the week distribution of highest ten ozone days for CAMS 3 & CAMS 38, % 20% Frequency of Days 15% 10% 5% CAMS 3 CAMS 38 0% Page 40 of 211

41 3.6 Frequency of High Ozone Days by Time of Day The time of day that peak ozone occurs is very consistent. Figure 3-16 shows the hour of peak 1-hour ozone concentrations for all of the area monitors during Region-wide, 96% of the peak times fall between 11am and 6pm. Reviewing the data for regulatory monitors only (CAMS 3 and 38) reveals the same pattern, with peaks at 1pm and 4pm, and more than 90% of all maximum 1-hour ozone concentrations occurring after noon. Assuming that this hour determines the likelihood of high ozone concentrations, however, would be incorrect. Instead, the key to this pattern is the correlation between these hours and important daytime meteorological conditions that are further discussed in Section 4. As the figure below shows, these results are consistent across the thresholds used to define a high ozone day. Figure Distribution of peak 1-hour ozone concentrations by hour for all Austin-Round Rock MSA monitors, % 20% 15% 10% >65 ppb >70 ppb >75 ppb 5% 0% When reviewing the patterns for the highest four and ten 8-hour ozone measurements per year at CAMS 3 and CAMS 38, the peak 1-hour ozone concentration is more frequently early afternoon for the highest concentrations in a given year, as shown in Figure 3-17 and Figure Although the overall shape of the curve is similar to the trends shown by all the monitors, the top four and top ten ozone levels appear to occur no later than 7pm. Page 41 of 211

42 Figure Hourly distribution of peak 1-hour ozone for highest four 8-hour concentrations, CAMS 3 & CAMS 38, % 20% 15% 10% 5% C3 C38 0% Hour Figure Hourly distribution of peak 1-hour ozone for highest ten 8-hour concentrations, CAMS 3 & CAMS 38, % 20% 15% 10% 5% C3 C38 0% Hour Since ozone concentrations are evaluated as an 8-hour average for regulatory purposes, it is also helpful to analyze which 8-hour ozone periods are most likely to record the peak daily 8-hour ozone concentrations on high ozone days. Most daily peak 8-hour ozone averages on high ozone days begin at 10 am or 11 am CST and end at 6 pm or 7 pm. This is consistent with the analysis above that shows the majority of peak 1-hour concentrations occurring between noon and 5 pm. Page 42 of 211

43 Figure Distribution of times for peak 8-hour ozone concentrations for all Austin-Round Rock MSA monitors, % 40% 30% 20% 10% >65 ppb >70 ppb >75 ppb 0% The top four and ten concentrations at the regulatory monitors, CAMS 3 and CAMS 38, exhibit the same pattern. Most peak ozone events occur in the middle of the day (between 10 am-7 pm), as expected. Although the two monitors are sited at different distances from the urban center, with CAMS 38 located approximately 15 miles northwest of the more centrally located CAMS 3, their temporal patterns of peak ozone are very similar. Figure Distribution of times for highest four 8-hour ozone concentrations, CAMS 3 and CAMS 38, % 50% 40% 30% 20% C3 C38 10% 0% Page 43 of 211

44 Figure Distribution of times for highest ten 8-hour ozone concentrations, CAMS 3 and CAMS 38, % 40% 30% 20% C3 C38 10% 0% 4 Local Meteorological Conditions Conditions conducive to the transport, formation, and accumulation of ozone are highly dependent on the prevailing large-scale weather patterns and the associated local meteorological conditions. Important local meteorological variables affecting ozone concentrations include solar radiation, clouds and precipitation, wind speed, wind direction, mixing height, and atmospheric stability. This section explores the relationship between daily peak ozone concentrations in Austin and available surface measurements of temperature, wind speed, wind direction, and relative humidity. Surface meteorological data from Murchison (CAMS 3) is shown in this analysis because this dataset provides a continuous record for and is assumed to be representative of conditions in the Austin-Round Rock area except where noted. For consistency and to provide a good regional average for comparison, temperature measurements were taken from the Austin-Bergstrom International Airport (ABIA). Since relative humidity is not measured at most monitoring sites, observations from ABIA meteorological monitors were also used for all relative humidity comparisons and is assumed to be representative of regional conditions. The table below shows a side by side comparison of the analyses performed in the 2012 conceptual model to those performed in this current conceptual model. For some of the parameters, CAPCOG went into further detail analyzing if there were more appropriate analyses and comparisons to better understand the relationships between high ozone and the meteorological conditions below. Page 44 of 211

45 Table 4-1. Comparison of 2012 and 2015 Conceptual Model Meteorological Analyses Meteorological Factor Temperature Wind Speed Relative Humidity Wind Direction 2012 Conceptual Model 2015 Conceptual Model Ozone vs. Max Temp. for C3 95 th Percentile Conditions for Days >= 75 ppb at C3 Ozone vs. Wind Speed for C3 95 th Percentile Conditions for Days >= 75 ppb at C3 Ozone vs. Average Daily Relative Humidity for C3 95 th Percentile for C3 Days >= 75 ppb grouped by Wind Direction for C3 and C38 Frequency of occurrence comparing all days and high ozone days at C3, controlling for temperature, wind speed, and relative humidity Ozone vs. Max Temp. for all monitors Ozone vs. Change in Temp. for all monitors 95 th and 100 th Percentiles for >75 ppb, > 70 ppb, >65 ppb, top 4, and top 10 days at C3 & C38 Ozone vs. Wind Speed for all monitors 95 th and 100 th Percentiles for >75 ppb, > 70 ppb, >65 ppb, top 4, and top 10 days at C3 & C38 Ozone vs. Average Daily Relative Humidity for all monitors Ozone vs. 2 pm Relative Humidity for all monitors 95 th and 100 th Percentiles for >75 ppb, > 70 ppb, >65 ppb, top 4, and top 10 days at C3 & C38 Days > 75 ppb, > 70 ppb, and > 65ppb grouped by wind direction for all monitors Direction frequency for high ozone days and all days and high ozone days at C3 & C38, controlling for temperature, wind speed, and relative humidity Direction frequency comparison of low and high ozone days at C3 & C38 For each meteorological parameter discussed below, using the 95 th percentile to describe typical conditions is preferred, as it eliminates outliers from the usual patterns. However, to the extent examining necessary conditions is valuable, CAPCOG has included the 100 th percentile as well to include outlier points. In some cases, the 100 th percentile for the meteorological conditions of highest four and highest ten ozone values are equal due to these outlying points. 4.1 Temperature The figure below presents a scatterplot of daily maximum 8-hour ozone concentrations vs. temperature for CAMS 3 (Murchison) from This figure shows a general but weak correlation between higher temperature and higher ozone concentrations. However, high daily peak temperature alone is not a sufficient condition for the formation of ozone. On a 100 o F or higher day, ozone concentrations can range from ppb. While there is a slight upward trend in ozone as peak temperatures increase, it is clear that the variability is too great to use temperature alone as an indicator of the potential for a high ozone day. Page 45 of 211

46 Nonetheless, high temperature is one important factor in creating the right conditions for higher ozone formation. 99% of ozone concentrations over 70ppb at Murchison occurred on days with a maximum temperature 85 o F or higher, and 88% over 65ppb. The average peak daily temperature on a day over 75ppb was 97.2 o F at the Murchison monitoring station. Figure 4-1. Maximum 8-hour ozone concentrations at CAMS 3 (ppb) vs. maximum daily temperature ( o F) at ABIA Ozone (ppb) y = 0.220x R² = Max Regional Temperature (deg F) An increase in the maximum ozone concentrations with warmer temperatures is not unexpected. Sunlight is necessary to drive the photochemical reactions leading to ozone formation, and warm temperatures are well-correlated with high amount of solar radiation. In addition, warmer temperatures may often be associated with overall meteorological conditions that are conducive to the formation and accumulation of ozone. For example, relatively warm temperatures occur preferentially during summer weather patterns conducive to large-scale stagnation. As evidence, similar patterns are observable across all the regional monitors, indicating that higher ozone levels tend to occur on days with higher peak temperatures. Table 4-2 shows the correlation for each monitor relative to maximum daily temperature, as well as for all monitors together. None of the monitors have an especially strong correlation to maximum temperature, evidenced by low R 2 values, but they all do show positive trends for ozone with increased temperature, with the exception of at CAMS 684, which had a nearly neutral trend. Page 46 of 211

47 Table 4-2. Ozone-maximum temperature correlation for all Austin-Round Rock MSA monitors, Monitor ppb/ o F R 2 Value Audubon (C3) Murchison (C38) Pflugerville (C613) Dripping Springs (C614) Round Rock (C674) McKinney Roughs (C684) Georgetown (C690) Gorzycki (C1603) Lockhart (C1604) San Marcos (C675+C1675) Hutto (C6602) Elroy Liberty Hill A more interesting comparison is to look at 8-hour ozone concentrations correlated to daily temperature changes. This positive relationship is much stronger, with R 2 values ranging from , showing that although there is still quite a bit of variability, a large temperature change over the course of the day improves the likelihood of higher ozone levels. This may explain why the highest ozone levels typically occur in shoulder seasons (in Central Texas, late spring or early fall), when daytime temperatures may be quite high while nighttime lows are dropping. This is consistent with the earlier analysis on frequency of high ozone days by month, which showed that the peaks in the season do indeed occur in May-June and August-September time frames. Figure 4-2. Maximum 8-hour ozone concentration at CAMS 3 (ppb) vs. diurnal differential temperature ( o F) at ABIA Ozone (ppb) y = 0.912x R² = Diurnal Differential Regional Temperature (deg F) Page 47 of 211

48 However, as with maximum daily temperature, a large difference in temperature is not a sufficient condition alone for the formation of ozone. Demonstrated in the plot above for CAMS 3, the largest temperature swing was 56 degrees, but the day measured a maximum 8-hour ozone concentration of only 45 ppb. Conversely, an ozone concentration of 81 ppb occurred on a day with only a 20 degree temperature change. For comparison, the 95 th and 100 th percentile values are presented below. 95% or 100% of values fall above the temperatures shown for different levels of ozone at CAMS 3 and CAMS 38. While the 95 th percentile provides a better representation of typical meteorological conditions for ozone formation, it is also interesting to examine the minimum temperatures observed for each ozone level, which could be considered the minimum necessary conditions for high ozone formation. For example, in nine years of data, 95% of ozone concentrations greater than 75 ppb at CAMS 38 occurred on days with peak temperatures over 92.3 o F, and 100% of the time ozone concentrations greater than 75 ppb at the same monitor occurred on days with peak temperatures over 88.0 o F. These conditions are summarized below. These tables also include the 95 th and 100 th percentile values for the annual top four and top ten ozone concentrations. Table th and 100 th Percentile for maximum daily temperature for CAMS 3 & CAMS 38, Ozone Level CAMS 3 95 th Percentile CAMS th Percentile CAMS th Percentile CAMS th Percentile >75 ppb 87.4 o F 81.0 o F 92.3 o F 88.0 o F >70 ppb 82.8 o F 78.1 o F 82.4 o F 81.0 o F >65 ppb 81.1 o F 75.9 o F 82.0 o F 75.9 o F Top o F 81.0 o F 82.8 o F 81.0 o F Top o F 75.9 o F 82.0 o F 75.9 o F All values 66.0 o F 40.0 o F 66.0 o F 40.0 o F Table th and 100 th Percentile for diurnal temperature change for CAMS 3 & CAMS 38, Ozone Level CAMS 3 95 th Percentile CAMS th Percentile Page 48 of 211 CAMS th Percentile CAMS th Percentile >75 ppb 24.9 o F 20.0 o F 24.1 o F 20.0 o F >70 ppb 23.8 o F 16.9 o F 24.4 o F 20.0 o F >65 ppb 23.0 o F 16.9 o F 23.3 o F 18.0 o F Top o F 24.8 o F 24.7 o F 20.0 o F Top o F 16.9 o F 23.0 o F 16.9 o F All values 12.1 o F 1.1 o F* 12.2 o F 1.1 o F *This data is from 3/13/2009 at ABIA, and temperature measurements from CAMS 3 & 38 both confirm that there was indeed only a 1 degree temperature change that day. 4.2 Wind Speed Much in the same way that high maximum daily temperatures and large changes in temperature correlate with higher ozone levels, low average daytime wind speeds typically correlate with higher ozone levels. All Austin-Round Rock MSA regional monitors display a negative relationship between maximum ozone and average daytime wind speeds. For this analysis, wind speeds are based on hourly observations collected at each ozone monitor between the hours of 6am and 6pm. The hourly wind

49 speeds, combined with corresponding wind directions, were converted to vectors for these hours, summed, and then determined an average magnitude. The highest ozone levels occurred on days with relatively low wind speeds. For CAMS 3, 95% of days with maximum ozone levels greater than 75ppb had wind speeds less than 5.4mph, and 100% of those days had wind speeds less than 6.5mph. At CAMS 38, 100% of days with ozone levels greater than 75ppb had wind speeds less than 5.0mph. Table 4-5 summarizes the 95 th percentile for the other ozone levels for CAMS 3 and 38. This table also presents the 95 th percentile for the top 4 and 10 days annually. Table th Percentile for wind speed for CAMS 3 & CAMS 38, Ozone Level CAMS 3 Wind Speed CAMS 38 Wind Speed >75 ppb 5.4 mph 4.8 mph >70 ppb 6.1 mph 5.5 mph >65 ppb 8.3 mph 7.4 mph Top mph 8.4 mph Top mph 7.2 mph All values 10.3 mph 9.3 mph Light wind speeds indicate conditions of limited horizontal ventilation and poor atmospheric dispersion, which are often associated with stable atmospheric conditions as well. As wind speed increases, improved horizontal ventilation and more efficient mixing throughout a deeper atmospheric boundary layer favor relatively lower ozone concentrations. In addition, higher wind speeds are sometimes associated with other meteorological conditions (such as rain showers or colder surface temperatures associated with strong fronts or low pressure systems) that are not conducive to the formation and/or accumulation of ozone. Figure 4-3. Maximum 8-hour ozone concentrations (ppb) vs. average wind speed (mph) between 6am-6pm at CAMS 3, Ozone (ppb) y = x R² = Wind Speed (mph) Page 49 of 211

50 There is a noticeable tendency for lower wind speeds to produce higher ozone, evidenced by the slight negative slope of the trendline. All monitors produced a similar trend when plotted against ozone concentrations, shown in the table below. Table 4-6. Ozone-wind speed correlation for all Austin-Round Rock MSA monitors, Monitor ppb/mph R 2 Value Audubon (C3) Murchison (C38) Pflugerville (C613) Dripping Springs (C614) Round Rock (C674) McKinney Roughs (C684) Georgetown (C690) Gorzycki (C1603) Lockhart (C1604) San Marcos (C675+C1675) Hutto (C6602) Elroy Liberty Hill While some monitors have slightly better correlations, none of the relationships between wind speed and maximum daily ozone can be considered determinative of ozone levels. Low daily wind speeds are not alone a sufficient condition for high ozone, made clear by the R 2 values shown above. Even at less than 2 mph, ozone concentrations can vary from ppb. For the two regulatory monitors, the maximum wind speeds observed for each of the high ozone thresholds are displayed in Table 4-7. To the extent that these wind speeds were the highest observed that coincides with peak ozone formation, wind speeds below these levels could be considered necessary for peak ozone formation. Table 4-7. Maximum wind speeds for CAMS 3 & CAMS 38, Ozone Level CAMS 3 Wind Speed CAMS 38 Wind Speed >75 ppb 6.5 mph 5.0 mph >70 ppb 8.8 mph 6.6 mph >65 ppb 15.4 mph 11.9 mph Top mph 9.0 mph Top mph 9.0 mph 4.3 Relative Humidity There are several ways to examine relative humidity s relationship to ozone formation discussed below. One option is to compare ozone to average relative humidity for the day, ideally encompassing any shifts or changes in that day s conditions. Alternatively, one can compare maximum ozone concentrations to the relative humidity from 2-3pm, right in the middle of the time of day ozone concentrations peak, as shown in Section 3.6, which may be more representative of the conditions immediately surrounding the formation of ozone. Both analyses are reviewed below. Page 50 of 211

51 As previously mentioned, all relative humidity measurements came from meteorological monitoring at the Austin-Bergstrom International Airport (ABIA) because humidity measurements are not available at most of the monitors, and for the ones where humidity data is available, it is only available for CAPCOG assumed that the ABIA measurements are representative of the area. However, relative humidity data at ABIA are not available from 3/5/2014 through the end of Figure 4-4 is a scatterplot of daily maximum ozone concentrations versus the average daily relative humidity measured at ABIA. 95% of high ozone days over 75ppb had a daily average relative humidity less than 73%. For days with ozone levels greater than 70ppb and 65 ppb, the odds change to 75% and 73% respectively. Figure 4-4. Maximum ozone concentrations (ppb) at CAMS 3 vs. daily average relative humidity (%) at ABIA, Max 8-Hour Daily Ozone (ppb) y = x R² = Daily Avg Relative Humidity (%) So while there is an obvious decreasing trend in the figure above, low daily average relative humidity is not a sufficient condition for high ozone. Although average relative humidity has better correlation than for some of the other parameters, the large number of outliers at higher relative humidities indicates that this is not necessarily a reliable indicator that there is a higher tendency to observe relatively higher ozone concentrations at lower relative humidity values. For example, the highest ozone concentrations for CAMS 3 range between 40-80% average relative humidity, and the maximum ozone concentration of 94 ppb at CAMS 3 fell on a day with 57% average relative humidity. Figure 4-5, however, appears to more closely follow a trend of high ozone levels at lower relative humidities. While there is still some uncertainty, the highest ozone concentrations occur below 40% humidity. Table 4-8 below compares the two methods. In all cases, the R 2 value for the relationship is similar between methods. Page 51 of 211

52 Figure 4-5. Maximum ozone concentrations (ppb) at CAMS 3 vs. daily 2pm relative humidity (%) at ABIA, Max 8-Hour Daily Ozone (ppb) y = x R² = Daily 2pm Relative Humidity (%) Table 4-8. Ozone-relative humidity correlation for all Austin-Round Rock MSA monitors, Monitor Average 2pm ppb/% R 2 Value ppb/% R 2 Value Audubon (C3) Murchison (C38) Pflugerville (C613) Dripping Springs (C614) Round Rock (C674) McKinney Roughs (C684) Georgetown (C690) Gorzycki (C1603) Lockhart (C1604) San Marcos (C675+C1675) Hutto (C6602) Elroy Liberty Hill Table 4-9 and Table 4-10 present a more striking comparison, showing the relative humidity measurements for each method, for which 95% of the ozone at the given levels falls below. These tables also include the 95 th percentile relative humidity measurements for the annual four highest and ten highest peak 8-hourozone averages. While average relative humidity on high ozone days is not appreciably different than the 95 th percentile across all days, 2 pm relative humidity for 95% of high ozone days does differ quite a bit when compared to all days, regardless of ozone concentration. Page 52 of 211

53 Table th Percentile for Average Daily Relative Humidity and 2 pm Relative Humidity for CAMS 3 Ozone Level Average Relative Humidity 2 pm Relative Humidity >75 ppb 73.4% 38.0% >70 ppb 75.3% 39.9% >65 ppb 72.8% 43.1% Top % 37.5% Top % 43.1% All values 87.6% 80.5% Table th Percentile for Average Daily Relative Humidity and 2 pm Relative Humidity for CAMS 38 Ozone Level Average Relative Humidity 2 pm Relative Humidity >75 ppb 77.3% 34.1% >70 ppb 71.0% 39.1% >65 ppb 74.4% 43.3% Top % 40.0% Top % 43.5% All values 87.6% 80.5% A closer look at the humidity conditions that could be considered necessary for high ozone in Central Texas (100% of high ozone days fall below these relative humidity values) shows a similarly stark difference between average and 2 pm relative humidity. 2 pm relative humidity measurements on high ozone days were always below 36-53%, while average daily relative humidity values all fall below 89%. The large range of average relative humidity measurements observed on high ozone days relative to the range of 2 pm relative humidity measurements suggest that mid-day relative humidity measurements is a better predictor of high ozone than the daily average is. Table th Percentile for Average Daily Relative Humidity and 2 pm Relative Humidity for CAMS 3 Ozone Level Average Relative Humidity 2 pm Relative Humidity >75 ppb 89.2% 40.5% >70 ppb 89.2% 45.8% >65 ppb 89.2% 53.1% Top % 40.5% Top % 53.1% Page 53 of 211

54 Table th Percentile for Average Daily Relative Humidity and 2 pm Relative Humidity for CAMS 38 Ozone Level Average Relative Humidity 2 pm Relative Humidity >75 ppb 89.2% 36.2% >70 ppb 89.2% 50.6% >65 ppb 89.2% 53.1% Top % 39.2% Top % 53.1% The 2 pm relative humidity data generally indicate that high ozone days occur predominantly during periods characterized by relatively low atmospheric moisture levels in the lower atmosphere at mid-day. As mentioned previously, high ozone days in Central Texas typically have a ridge of high pressure that extends south or southwest into Texas. The large-scale clockwise circulation around this ridge of high pressure is often associated with the transport of air of recent continental origin into Austin from geographic areas located to the north and/or east of Central Texas. This air mass is often characterized by relatively low atmospheric moisture values and high amounts of background ozone, as opposed to the relatively moister and cleaner air masses that originate over maritime regions such as the Gulf of Mexico during periods of southerly flow. In addition, days with relatively high relative humidity values may indicate other large-scale meteorological conditions (such as cloudy and/or rainy days that sometimes occur in association with low pressure systems and flow from the Gulf) that are not conducive to the formation and/or accumulation of high ozone concentrations near the surface. 4.4 Wind Direction Wind direction is the most site-specific of the meteorological comparisons, as each monitor is likely to experience high ozone from different directions, depending on its location. Figure 4-6 and Figure 4-7 show that all of the monitors in the Austin-Round Rock MSA have different wind direction patterns that tend to align with high ozone concentrations. For example, while many of the high ozone days at CAMS 3 and CAMS 38 occur when winds are coming from the south-southeast, CAMS 3 tends to have more high ozone days when winds are coming from the northeast than CAMS 38 does, presumably because such winds would cross a much more heavily urbanized area to reach CAMS 3 than they would for CAMS 38. In the figures below, the depth of the bars indicate the frequency of high ozone day occurrence from a particular wind direction when the maximum 8-hour ozone concentration is greater than 65 ppb, 70, ppb, and 75 ppb. For this analysis, wind directions are based on hourly resultant wind direction observations collected at each ozone monitor between the hours of 6am and 6pm for The hourly wind directions, combined with corresponding wind speeds, were converted to vectors for these hours, summed, and then determined a cumulative direction. CAMS 1603 did not have any ozone measurements over 65ppb and is not included in the figures below. For most monitors, high ozone days occurred most often when the daytime winds had a northnortheasterly clockwise through south-southeasterly direction. This is consistent with previous models. Page 54 of 211

55 Figure 4-6. Resultant wind direction 6am 6 pm on high ozone days for CAMS 3, 38, 614, 674, 690, and , Page 55 of 211

56 Figure 4-7. Resultant wind direction 6am 6 pm on high ozone days at CAMS 613, 684, 6602, 1604, Liberty Hill, and Elroy, Page 56 of 211

57 To compare the frequency of occurrence of wind direction between low and high ozone days, Figure 4-8 through 4-13 present the frequency of wind direction as a percentage of total days grouped by all days and high ozone days at CAMS 3 and 38. To help isolate the importance of wind direction on days generally favorable to near-surface ozone formation in Central Texas, only days with criteria identified previously in this section as being the 95 th percentile values for temperature differential, 2pm relative humidity, and wind speed, but below the ozone threshold being analyzed (all else being equal). These figures are summarized below. Figure 4-8. Comparison of wind direction on days with ozone >75ppb to days with ozone <=75ppb, controlling for meteorological conditions at CAMS 3 WNW NW NNW 25% 20% 15% 10% N NNE NE ENE 5% W 0% E WSW ESE SW SE SSW S SSE Ozone >75ppb Ozone <=75ppb Figure 4-9. Comparison of wind direction on days with ozone >70 ppb to days with ozone <=70ppb, controlling for meteorological conditions at CAMS 3 WNW NW NNW 25% 20% 15% 10% N NNE NE ENE 5% W 0% E WSW ESE SW SE SSW S SSE Ozone >70ppb Ozone <=70ppb Page 57 of 211

58 Figure Comparison of wind direction on days with ozone >65ppb to days with ozone <=65ppb, controlling for meteorological conditions at CAMS 3 NNW 40% N NNE NW 30% NE WNW W 20% 10% 0% ENE E WSW ESE SW SE SSW Ozone >65ppb S SSE Ozone <=65ppb Table th Percentile for Average Daily Relative Humidity and 2pm Relative Humidity for CAMS 3 Ozone Level Greater Than Daily Temperature Change Less Than Wind Speed Less Than 2pm Relative Humidity >75 ppb 24.9 o F 5.4 mph 38.0% >70 ppb 23.8 o F 6.1 mph 39.9% >65 ppb 23.0 o F 8.3 mph 43.1% Figure Comparison of wind direction on days with ozone >75ppb to days with ozone <=75ppb, controlling for meteorological conditions at CAMS 38 WNW W NW NNW 30% 25% 20% 15% 10% 5% 0% N NNE NE ENE E WSW ESE SW SE SSW SSE S Ozone >75ppb Ozone <=75ppb Page 58 of 211

59 Figure Comparison of wind direction on days with ozone >70ppb to days with ozone <=70ppb, controlling for meteorological conditions at CAMS 38 NNW 20% N NNE NW 15% NE WNW W 10% 5% 0% ENE E WSW ESE SW SE SSW Ozone >70ppb S SSE Ozone <=70ppb Figure 4-13 Comparison of wind direction on days with ozone >65ppb to days with ozone <=65ppb, controlling for meteorological conditions at CAMS 38 WNW W NW NNW 30% 25% 20% 15% 10% 5% 0% N NNE NE ENE E WSW ESE SW SE SSW S SSE Ozone >65ppb Ozone <=65ppb Table th Percentile for Average Daily Relative Humidity and 2pm Relative Humidity for CAMS 38 Ozone Level Greater Than Daily Temperature Change Less Than Wind Speed Page 59 of 211 Less Than 2 pm Relative Humidity >75 ppb 24.1 o F 4.8 mph 34.1% >70 ppb 24.4 o F 5.5 mph 39.1% >65 ppb 23.3 o F 7.4 mph 43.3%

60 Finally, Figure 4-14 and 4-15 present higher ozone days compared to all low ozone days (ozone concentrations less than 55 ppb) for CAMS 3 and 38. Figure Comparison of wind direction on high ozone days to low ozone days at CAMS 3 WNW W NW NNW 30% 25% 20% 15% 10% 5% 0% N NNE NE ENE E Direction % when Ozone >75 Direction % when Ozone >70 Direction % when Ozone >65 WSW ESE Direction % when Ozone <55 SW SE SSW S SSE Figure Comparison of wind direction on high ozone days to low ozone days at CAMS 38 NNW 30% N NNE WNW W WSW NW SW 20% 10% 0% NE SE ENE E ESE Direction % when Ozone >75 Direction % when Ozone >70 Direction % when Ozone >65 Direction % when Ozone <55 SSW S SSE Page 60 of 211

61 This analysis shows that high ozone days occur more frequently from certain wind directions (northnortheasterly or south-southeasterly direction for CAMS 3, and easterly clockwise to southerly for CAMS 38) even when controlling for the other meteorological parameters that are conducive to ozone formation. What these charts also show, is that peak temperature, the difference between high and low temperatures, relative humidity, and wind speed may not be sufficient to explain high ozone formation without accounting for wind direction. Refer to Section 7 of this report for additional analysis of the local-scale transport paths prior to high ozone days at Austin monitoring locations during Background and Locally-Formed Ozone This section investigates the importance of ozone transported into the region on maximum ozone concentrations measured in the Austin-Round Rock MSA, and the typical local contribution to peak ozone formation. CAPCOG used the maximum and minimum peak 8-hour ozone measurements collected at regional ozone monitoring stations to estimate the amount of background and locallygenerated ozone on high ozone days. Previous analyses in former conceptual models for Central Texas investigated the levels of ozone transported into the Austin area and estimated the amount of ozone formed from local emissions sources. In the most recent version of the conceptual model, regional monitoring data from was used. 26 The analysis assumed that the lowest daily peak 8-hour ozone concentration represented the background ozone being transported into the area. UT then calculated the difference between the minimum and maximum daily peak 8-hour ozone concentrations to provide an estimate of the ozone increase associated with emissions from local sources. For this analysis, high ozone days were defined as those that had maximum ozone concentrations greater than 70 ppb at any monitor within the Austin- Round Rock MSA. Whereas previously, UT used a few nearby monitoring stations that were not located within the MSA to conduct this type of analysis, for this conceptual model, CAPCOG only used 8-hour ozone concentrations recorded within the MSA. It is important to note that these analyses are only intended to provide a reasonable estimate of background and locally-formed ozone using the available surface monitoring data and a simple methodology. This methodology implicitly assumes that the appropriate upwind and downwind ozone concentrations are successfully measured by the existing network of monitoring stations. The actual maximum and minimum ozone concentrations in the Austin area may have been substantially different than those measured by the existing monitoring network. This analysis also inherently assumes steadystate conditions in emissions, chemistry, meteorology, and background ozone throughout the Austin area so that the difference in minimum and maximum concentrations defines the impact of local emissions sources on ozone concentrations. Many of the monitors are likely impacted by emissions from MSA sources regardless of the prevailing surface wind direction. Additionally, for the analysis described in this section, no attempt was made to confirm that the monitor with the minimum concentration was indeed located on the upwind side of the urban area on any particular day. To the extent that the monitoring network in the later part of the period analyzed ( ) provides better spatial coverage than the network did in the earlier part of the period, there may be some bias in these results, 26 Page 61 of 211

62 but that may be counteracted by the significant decrease in emissions that occurred over this timeframe. The table below shows the average differences between the region s highest and lowest peak 8-hour ozone concentrations by month, as an example, using days when the peak ozone concentrations were greater than 70 ppb. This analysis indicates that transported ozone is estimated to have contributed 70-90% to the maximum concentrations over this period, with the average monthly background ozone contribution about 60 ppb. The amount of locally formed ozone, ranged from 2 to 41 ppb, depending on the event. The average for all days analyzed was about 16.5 ppb. Table 5-1. Monthly and total average statistics on >70ppb days based on the Austin-Round Rock MSA minimum and maximum ozone concentrations, Month # Events Avg Max Ozone (ppb) Avg Min Ozone (ppb) Avg Min/Max (%) Avg Delta (ppb) Min Delta (ppb) Max Delta (ppb) March % April % May % June % July % August % September % October % % The figure below presents the ratio of the lowest peak 8-hour ozone average to the highest peak 8-hour ozone average on days on which the maximum ozone concentrations were greater than 65 ppb, 70 ppb, and 75 ppb, alongside the estimates performed by UT on monitoring data. The relative contribution of background ozone in the current analysis is generally lower than the previous analysis, although this does not necessarily indicate a greater relative contribution from local sources in recent years. Since the analysis, the regional ozone monitoring network has expanded, especially upwind of the core urban area. This expansion has enabled a better characterization of upwind regional ozone concentrations, which in turn explains the larger differences between peak 8-hour ozone concentrations in the current analysis than what was reflected in the analysis. Page 62 of 211

63 Figure 5-1. Average minimum/maximum ozone concentrations for different high ozone levels in the Austin-Round Rock MSA, , compared to analysis 100% 90% Avg Min/Max Ozone 80% 70% 60% 50% 40% 30% 20% 10% Analysis, >=70ppb >65ppb >70ppb >75ppb 0% Background ozone concentrations do typically contribute more to peak 8-hour ozone averages on higher ozone days than it does on days when the region-wide peak 8-hour ozone concentration was below 60 ppb, as shown in Figure 5-2. Page 63 of 211

64 Figure 5-2. Average minimum/maximum ozone concentrations for different high ozone levels in the Austin-Round Rock MSA, compared to lower ozone days, % 90% Avg Min/Max Ozone 80% 70% 60% 50% 40% 30% 20% 10% 0% All Days <60 ppb >65ppb >70ppb >75ppb Overall, the results of CAPCOG s analyses shows that transport contributes substantially (74-94%) to the maximum 8-hour ozone concentrations measured on high ozone days in the Austin-Round Rock area. The comparison of CAPCOG s results for relative to UT s analysis for may reflect the expansion of the ozone monitoring network in recent years. Finally, unlike UT s analysis, CAPCOG s analysis shows a seasonal pattern in the relative contribution of background and local ozone levels, with a higher local contribution in the middle of the season, but a relatively flat background contribution of about 60 ppb. Further discussion on regional transport and back-trajectories is present in Sections 6 and 7. 6 Regional Transport Patterns Prior to High Ozone Days This section presents back-trajectory maps generated for high ozone days in Austin to visually summarize both the synoptic-scale long-range flow patterns that have occurred immediately prior and during high ozone events, and the most frequent geographic areas upwind of Central Texas corresponding to potential source regions of background ozone entering Austin. The regional backtrajectory maps were based on HYSPLIT back-trajectories initiated at heights above the surface (AGL) of 50 meters and 1 kilometer (1000 meters). Back-trajectories calculated using surface winds were used to estimate near-surface flow patterns and potential emissions source areas in the local Austin-Round Rock area. Maps were generated to investigate the spatial variability in transport patterns for high ozone days grouped by: (1) seasonal period, and (2) interstate, intra-state, and local geographic areas. All maps show back-trajectories for high ozone days that occurred during years The use of an Page 64 of 211

65 ensemble of back-trajectories provides a more robust analysis than that obtained with individual backtrajectories. The reader is referred to Section 2.0 of this report for information on the data and methodology used to calculate the back-trajectories. Section 2.0 also provides a discussion of uncertainties associated with the back-trajectory results. Throughout this section: 1) High ozone days are those with maximum 8-hour ozone concentrations >= 65 ppb at one or more Austin monitors, except for 2013 and 2014, when there were comparatively fewer days with ozone >= 65 ppb. For these two years the top ten ozone days for each year were included in the analysis. 2) Results for high ozone days are grouped into two periods: early season corresponding to April-July, and late season corresponding to August-October. 6.1 Regional Transport at CAMS 3 The following back-trajectory maps are based on 48-hour HYSPLIT back-trajectories initiated at a height of 50 m above the surface at the hour which recorded the highest ozone for that day. The five-day trajectory duration was selected to capture long-range transport not only within Texas, but also from distant areas, such as the central and southeastern US, during conditions of relatively low wind speeds in the lower atmosphere. The back-trajectory initialization height of 50 m AGL was used to represent wind flow near the surface on high ozone days in Central Texas. Figure 6-1 shows the 50m 48-hour back trajectories on high ozone days for every ozone season between 2006 and The map indicates that there are a variety of source regions of high ozone transport into the Austin area, ranging from northerly clockwise to southerly. There are relatively few back trajectories that show transport from west of Austin on high ozone days. The maps for the early (April July) and late (August October) seasons are shown in Figure 6-2 (a) and (b), respectively. During the early season, there remains great variety in the locations of upwind areas. Approximately half of the days show long-range transport from the northeast or north into Texas, while the other half of the days show transport from the southeast, well into the Gulf of Mexico. For the late season, long-range transport patterns indicate states to the north and northeast are typically in the upwind area. There is a subset of back trajectories that originates to the southeast of Austin near the Gulf coast, but these are shorter than those for the early season, indicating more stagnated conditions. Within Texas, upwind areas during the late season are primarily located to the east and northeast of Austin. Page 65 of 211

66 Figure m HYSPLIT 48-hour back trajectories originating at CAMS 3 (Austin Northwest), Page 66 of 211

67 Figure 6-2. (a) 50m HYSPLIT 48-hour back trajectories originating at CAMS 3 during April-July ( ); (b) 50m HYSPLIT 48-hour back trajectories originating at CAMS 3 during August-October ( ) (a) (b) The following back-trajectory maps are based on 48-hour HYSPLIT back-trajectories initiated at a height of 1000 m above the surface at the hour which recorded the highest ozone for that day. Because there is vertical transport of ozone in addition to horizontal transport, it is useful to analyze back trajectories at a higher altitude. Furthermore, wind flow at the surface can be quite different from wind flow above the surface. 1000m AGL was chosen as a secondary starting height because it is within the mixed layer during the time of peak ozone. Figure 6-3 shows the 1000m 48-hour back trajectories on high ozone days for every ozone season between 2006 and Similar to the 50m back trajectory analysis, the 1000m back trajectories were then split between early season (April July) and late season (August October). It is apparent that the 1000m back trajectories are generally longer than those initiated at 50m. This is due to the absence of surface friction at upper levels. Figure 6-4(a) and (b) show 48-hour 1000m back trajectories during the early and late parts of ozone season, respectively. The early ozone season appears to have a large variety of source regions from all directions, while the late season back trajectories are predominantly out of the north and northeast. Page 67 of 211

68 Figure m HYSPLIT 48-hour back trajectories originating at CAMS 3, Page 68 of 211

69 Figure 6-4. (a) 1000m HYSPLIT 48-hour back trajectories originating at CAMS 3 during April-July ( ); (b) 1000m HYSPLIT 48-hour back trajectories originating at CAMS 3 during August-October ( ) (a) (b) The distance and direction from CAMS 3 of each hourly back trajectory point was calculated in order to quantify the most common source regions on high ozone days. These hourly back trajectory points for the 50m level are shown in Figure 6-5. For this analysis, the region of central Texas within a 500 km radius of CAMS 3 was partitioned into octants: northern, northeastern, eastern, southeastern, etc. The region was further subdivided by distance boundaries: areas within 100 km of C3, 100 to 200 km of C3, etc., out to 500 km from C3 and beyond. Table 6-1 contains the number of hourly air parcel positions within each sub-division and the total for each octant and distance. The total for each distance subdivision of these octants will be referred to as bin counts. When looking at the ozone season as a whole, it is difficult to determine a dominant source region of surface air flow. Table 6-1 indicates that 34.3% of the bin counts were in the S or SE octants, while 40.4% of the bin counts were in the N or NE octants. It also shows that northerly flow was more associated with longer-range transport, with 21.7% of the northerly bin counts outside of the 500 km range. Southerly source regions were associated with shorter-range transport, with only 7.9% of the southerly bin counts outside the 500 km range. Page 69 of 211

70 Figure m HYSPLIT 48-hour back trajectories originating at CAMS 3, Page 70 of 211

71 Table 6-1. Bin counts for hourly 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 3, m Bin Counts Direction km km 300km 400km 500km 500+ km Total N NE E SE S SW W NW Total 2,774 2,266 1, Figure 6-6. (a) 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 3 during April-July ( ); (b) 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 3 during August-October ( ) A bin count analysis was also conducted for the April July ozone season and the August October ozone season for both 50m and 1000m HYSPLIT back trajectories originating at CAMS 3. Figure 6-6(a) and (b) show the hourly back trajectory points at the 50m level for high ozone days during the early and late ozone seasons, respectively. For the early ozone season, there appears to be a fairly even distribution of directional source regions of high ozone at the 50m level. Table 6-2reveals that southerly source regions close to the surface tend to be the most common during the early ozone season, with 25.9% of the bin counts occurring in the S octant. The second most common source regions are SE, NE, and N with between 14.9% and 15.7% of the bin counts in each of these three octants. The bin count Page 71 of 211

72 analysis for August October high ozone days (Table 6-3) shows major differences in transport patterns compared to the early ozone season. The late ozone season bin counts indicate that the majority of high ozone transport occurs from the north and northeast (50.7%). There are few back trajectory points west of Austin during the late season, in contrast to the April July season. Back trajectory points beyond 500km were less common during the late ozone season compared to the early ozone season. Table 6-2. Bin counts for hourly 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 3 for April-July ( ) Direction km km 50m Bin Counts (April-July) km km km 500+ km Total N NE E SE S SW W NW Total Table 6-3. Bin counts for hourly 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 3 for August-October ( ) 50m Bin Counts (August-October) Direction km km 300km 400km 500km 500+ km Total N NE E SE S SW W NW Total At the 1000m level (Figure 6-7), there is a clearer trend of northerly source regions, although there is still great variety in the source directions. The bin count analysis in Table 6-4 supports this, showing 44.1% of the bin counts in the N or NE octants and only 24.8% in the S or SE octants. Combined, these two sets of octants account for a lesser share of the total bin counts at the 1000m level than at the 50m level. As expected due to frictional effects at the surface, bin counts outside the 500km range were far more numerous for 1000m back trajectories (25.0%) than for 50m back trajectories (8.6%). Page 72 of 211

73 Figure m HYSPLIT 48-hour back trajectory hourly points ending at CAMS 3, Page 73 of 211

74 Table 6-4. Bin counts for hourly 1000m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 3, m Bin Counts Direction km km 300km 400km 500km 500+ km Total N NE E SE S SW W NW Total Figure 6-8(a) and (b) show the hourly back trajectory points at the 1000m level for high ozone days during April July and August October. There appears to be a fairly even distribution of directional source regions of high ozone during the early season. Table 6-5reveals that easterly source regions tend to be the most common during the early ozone season, although the results were fairly evenly split between NE, E, SE, and S with between 16.6% and 18.1% of the bin counts occurring in each of those four octants. Table 6-6 shows that for the late ozone season, 60.1% of the bin counts occurred in the NE octant. Figure 6-8. (a) 1000m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 3 during April-July ( ); (b) 1000m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 3 during August-October ( ) (a) (b) Page 74 of 211

75 Table 6-5. Bin counts for hourly 1000m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 3 for April-July ( ) 1000m Bin Counts (April-July) Direction km km 300km 400km 500km 500+ km Total N NE E SE S SW W NW Total Table 6-6. Bin counts for hourly 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 3 for August-October ( ) 1000m Bin Counts (August-October) Direction km km 300km 400km 500km 500+ km Total N NE E SE S SW W NW Total Regional Transport at CAMS 38 The following back-trajectory maps are based on 48-hour HYSPLIT back-trajectories initiated at a height of 50 m above the surface at the hour which recorded the highest ozone for that day. The backtrajectory initialization height of 50 m AGL was used to represent wind flow near the surface on high ozone days in Central Texas. These back trajectories originated from CAMS 38 (Audubon) and are used to determine if different transport patterns exist for that station compared to CAMS 3 (Austin Northwest). Figure 6-9 shows the 50m 48-hour back trajectories on high ozone days for every ozone season between 2006 and The map indicates that while there are a variety of source regions of high ozone transport into the Austin area, there are even fewer back trajectories that show transport from west of Austin on high ozone days compared to CAMS 3. The maps for the early (April July) and late (August October) seasons are shown in Figures 6-10 (a) and (b), respectively. It is difficult to discern any additional differences in transport patterns between CAMS 3 and CAMS 38, especially during the late ozone season. Page 75 of 211

76 Figure m HYSPLIT 48-hour back trajectories originating at CAMS 38, Page 76 of 211

77 Figure (a) 50m HYSPLIT 48-hour back trajectories originating at CAMS 38 during April-July ; (b) August- October, (a) (b) The following back-trajectory maps are based on 48-hour HYSPLIT back-trajectories initiated at a height of 1000 m above the surface at the hour which recorded the highest ozone for that day. Figure 6-11 shows the 1000m 48-hour back trajectories on high ozone days for every ozone season between 2006 and As with the 50 m back trajectories, the 1000 m back trajectories for CAMS 38 appear to have fewer cases of transport from the northwest when compared to CAMS 3. Similar to the 50m back trajectory analysis, the 1000m back trajectories were then split between early season (April July) and late season (August October). Figures 6-4 (a) and (b) show 48-hour 1000m back trajectories during the early and late parts of ozone season, respectively. The early ozone season appears to have a large variety of source regions from all directions, while the late season back trajectories are predominantly out of the north and northeast. This is a similar transport pattern as seen with CAMS 3. Page 77 of 211

78 Figure m HYSPLIT 48-hour back trajectories originating at CAMS 38, Page 78 of 211

79 Figure m HYSPLIT 48-hour back trajectories originating at CAMS 38 during April-July ; (b) August- October, (a) (b) Page 79 of 211

80 Figure m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 38, Page 80 of 211

81 The following bin count analysis focuses on the differences in source regions between CAMS 3 and CAMS 38. The 50 m bin counts for the entire season, presented in Table 6-7, shows that transport from the NE clockwise to S were more common for CAMS 38 than for CAMS 3. Northerly transport was less common for CAMS 38 (11.2%) than for CAMS 3 (18.0%). Bin counts for the early ozone season (Table 6-8) at 50 m showed similar results, except that NE transport was more common for CAMS 3 by 0.5%. However, this was not the case for the late ozone season (Table 6-9). Northeasterly transport was more common by 3% for CAMS 38 than for CAMS 3. Table 6-7. Bin counts for hourly 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 38, m Bin Counts Direction km km 300km 400km 500km 500+ km Total N NE E SE S SW W NW Total Figure (a) 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 3 during April-July ( ); (b) 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 38 during August-October ( ) (a) (b) Page 81 of 211

82 Table 6-8. Bin counts for hourly 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 38 for April-July ( ) Direction km km 50m Bin Counts (April-July) km km km 500+ km Total N NE E SE S SW W NW Total Table 6-9. Bin counts for hourly 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 38 for August- October ( ) 50m Bin Counts (August-October) Direction km km 300km 400km 500km 500+ km Total N NE E SE S SW W NW Total Page 82 of 211

83 Figure m HYSPLIT 48-hour back trajectory hourly points ending at CAMS 38, Page 83 of 211

84 Table Bin counts for hourly 1000m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 38, m Bin Counts Direction km km 300km 400km 500km 500+ km Total N NE E SE S SW W NW Total An analysis of 1000 m back trajectory bin counts for CAMS 38 yielded similar results as the 50 m analysis when compared to CAMS 3. Longer range transport is more common at 1000 m than for 50 m and both stations have a similar percentage of bin counts > 500 km (25.0% for CAMS 3 and 23.7% at CAMS 38). Bin counts from NE clockwise to S source regions were more common for CAMS 38 for the entire season, but for the early season (Table 6-11), NE bin counts were more common for CAMS 3. For the late ozone season at CAMS 38 (Table 6-12), 58.5% of bin counts were in the N and NE octants, which is actually less than the number of N or NE bin counts for CAMS 3. Figure (a) 1000m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 38 during April-July ( ); (b) 1000m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 38 during August-October ( ) (a) (b) Page 84 of 211

85 Table Bin counts for hourly 1000m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 38 for April-July ( ) 1000m Bin Counts (April-July) Direction km km 300km 400km 500km 500+ km Total N NE E SE S SW W NW Total Table Bin counts for hourly 50m HYSPLIT 48-hour back trajectory hourly points originating at CAMS 38 for August- October ( ) 1000m Bin Counts (August-October) Direction km km 300km 400km 500km 500+ km Total N NE E SE S SW W NW Total Large-Scale Weather Patterns during High Ozone Episodes The conditions conducive to the transport, formation, and accumulation of ozone in Central Texas are primarily dependent on the prevailing large-scale weather patterns. AACOG investigated the large-scale atmospheric circulation features during high ozone episodes in the Austin area for using upper air and surface weather maps maintained by the Precipitation Diagnostics Group in the Mesoscale and Microscale Meteorology Division of NCAR. 27 For the purposes of these large-scale weather analyses, high ozone episodes were defined as multi-day periods having at least one day with 8-hour ozone concentrations >= 70 ppb at one or more Austin monitors. Since the observational analysis of weather features is time and resource intensive, the threshold concentration of 70 ppb was chosen to restrain the number of reviewed episodes to a reasonable, but representative, number. The dates of the twelve high ozone episodes are shown in Table 8-1. AACOG developed five-day HYSPLIT back-trajectories for days with 8-hour ozone concentrations of at least 70 ppb Page 85 of 211

86 Table high ozone episodes reviewed for the Austin area Episode Number Episode Dates Days with 8-Hr O3 >= 70 ppb Austin Area Daily Maximum 8-Hr Ozone (ppb) 1 May 13 23, , 70, 58, 70, 78, 67, 54, 53, 73, 68, 44 2 May 30 June 2, , 61, 72, 44 3 June 22 28, , 68, 63, 75, 87, 81, 61 4 Aug 9 12, , 81, 94, 52 5 Aug 18 22, , 44, 81, 72, 66 6 Sep 8 12, , 65, 77, 64, 56 7 May 10 14, , 64, 60, 70, 56 8 June 1 9, , 58, 79, 71, 56, 64, 67, 71, 54 9 July 1 7, , 67, 82, 75, 78, 68, Aug 16 20, , 70, 70, 66, Sep 22 27, , 73, 72, 89, 69, May 15 18, , 70, 66, 58 Table 8-2 presents the results of the large-scale weather analyses. Descriptive discussions of the relevant upper air and surface large-scale weather features are provided in addition to summaries of local meteorological observations at Austin Northwest (maximum daily temperature: T, average morning (6a 10a) and afternoon (12p 4p) wind speed: WS, morning and afternoon wind direction: WD). Wind directions and wind speeds were broken down into morning and afternoon components in order to capture any directional wind shifts that may have occurred during the day, as well as to distinguish days with directional wind shifts from those that had more stagnant air flow. The daily average relative humidity (RH) was obtained using data from TCEQ at Austin-Bergstrom International Airport for 2012 and 2013, and Burnet County Airport for Also provided in Table 8-2 are the minimum and maximum 8-hour daily maximum ozone concentrations measured by the Austin area monitoring network. To better capture the development and evolution of the weather patterns throughout each of the high ozone episodes, AACOG included one or more days prior and following the high ozone days. Table rows are color-coded according to the maximum ozone concentration: Blue: ppb; Green: ppb; and Yellow: 75 ppb or higher. Site abbreviations are as follows: NW: Austin Northwest (CAMS 3); A: Audubon (CAMS 38); DS: Dripping Springs (CAMS 614); MR: McKinney Roughs (CAMS 684); LG: Lake Georgetown (CAMS 690); G: Gorzycki (CAMS 1603); L: Lockhart (CAMS 1604); SM: San Marcos (CAMS 1675); and H: Hutto (CAMS 6602). Page 86 of 211

87 Table 7-2. Local meteorological measurements at Austin Northwest and large-scale weather features for high ozone episodes in Austin during Episode and Date Episode 1: 5/13/2012 Episode 1: 5/14/2012 Episode 1: 5/15/2012 Max Austin 8- Hr Ozone (ppb) and Site Min Austin 8-Hr Ozone (ppb) and Site Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD 62 (SM) 54 (LG) N 4.9 N 70 (DS) 60 (LG) N 3.5 NNE 58 (SM) 50 (LG) NE 2.6 WNW Discussion of Selected Large-Scale Weather Features Surface high pressure system builds over the Great Plains and extends into central TX in the wake of a cold front, producing northerly surface winds throughout the day. An upper level trough sits over the middle of the country, but moderates to a more zonal flow southward to central TX. Mid-level low pressure develops over Eastern New Mexico early in the day and moves southeast, approaching central Texas by the end of the day. Pressures fall throughout the day. Clouds and precipitation develop later in the afternoon over central Texas. Surface winds continue out of the north with high pressure over the Rockies. Surface winds are light and variable in the morning. Low pressure system clears central Texas after noon, reestablishing northerly surface flow. Stronger northerly winds give way to calmer conditions after the low moves away from the area. Surface high pressure continues to dominate the central plains. Page 87 of 211

88 Episode and Date Episode 1: 5/16/2012 Episode 1: 5/17/2012 Episode 1: 5/18/2012 Episode 1: 5/19/2012 Max Austin 8- Hr Ozone (ppb) and Site Min Austin 8-Hr Ozone (ppb) and Site Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD 70 (A) 58 (MR) NNW 2.8 NNE 78 (A/LG) 63 (MR) NNW 0.7 NE 67 (A) 54 (DS) 61 (NW/H) 51 (NW/M R/LG) S 1.8 SSE S 1.5 SSE Discussion of Selected Large-Scale Weather Features Surface high pressure builds over SW Texas resulting in slightly less humid conditions and northerly flow. In the upper levels, a trough extends from the Ohio Valley into east Texas and moves east. Base of the trough is tilted such that winds aloft shift from northerly to westerly over central TX. Large surface high pressure system extends from the Great Lakes, across the southeast U.S. and into Texas providing more stagnant air flow. Northerly winds in the morning give way to generally easterly winds later in the day. Upper level trough moves into the southeastern U.S. Upper level ridge begins to move into central Texas from the west. Weak steering in the mid levels. Weak SE surface winds begin to increase in the afternoon with high pressure over the southeast U.S. High pressure persists over southeast U.S. SE surface flow with a frontal boundary moving across west Texas. Upper level ridge prevents storm formation along the boundary. Page 88 of 211

89 Episode and Date Episode 1: 5/20/2012 Episode 1: 5/21/2012 Episode 1: 5/22/2012 Episode 1: 5/23/2012 Max Austin 8- Hr Ozone (ppb) and Site Min Austin 8-Hr Ozone (ppb) and Site Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD 53 (H) 50 (DS) ESE 1.5 NE 73 (LG) 57 (MR) NNW 1.3 NNE 68 (DS) 63 (MR) NW 0.7 SSE 44 (DS) 39 (MR/H) S 2.6 SSE Discussion of Selected Large-Scale Weather Features A cool front slowly makes its way through north TX. Surface winds in the Austin area are stagnant for much of the day with a slight easterly component. Upper level winds are also light over central TX with a large area of high pressure over Texas and northern Mexico. Central TX is located at the tail end of the frontal boundary, bringing a brief shift in surface flow from easterly to northerly. Texas, northern Mexico, and the Southwest continue to be under the influence of a large upper level high pressure system. Morning surface winds are northwesterly but shift to southeasterly in the afternoon as high pressure moves into the southeastern U.S. in the wake of the front. Upper level ridge over northern Mexico causes NW flow aloft. Southerly surface winds continue as high pressure over the southeastern U.S. continues to bring in clean maritime air into central TX. Upper level flow continues to be northwesterly. Page 89 of 211

90 Episode and Date Episode 2: 5/30/2012 Episode 2: 5/31/2012 Episode 2: 6/1/2012 Episode 2: 6/2/2012 Max Austin 8- Hr Ozone (ppb) and Site Min Austin 8-Hr Ozone (ppb) and Site Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD 39 (DS) 32 (MR) SSE 1.7 ESE 61 (A) 47 (SM) E 1.7 NE 72 (A) 61 (SM) NNE 1.3 NE 44 (LG) 35 (MR) S 1.6 SSE Discussion of Selected Large-Scale Weather Features Frontal boundary about to enter north TX. Mesoscale convective complex moves along the boundary, causing thunderstorms across eastern OK. SE surface winds occur ahead of the front in central TX. Large upper level ridge sits over southern Baja Peninsula, bringing NW winds aloft. Cool front pushes through central TX in the afternoon, causing a wind shift from SE to NE and a trace of rain. In the upper levels, the trough associated with the frontal boundary becomes more pronounced and the ridge persists across northern Mexico and west TX. Cool front clears Texas. Surface winds are NE shifting to E throughout the day as high pressure builds over the Great Plains. Another upper level trough of low pressure develops over the SW U.S., oriented E-W, which develops into a cut-off low over northern Mexico. Upper level low persists over northern Mexico. Surface winds are now out of the SW to SE due to high pressure moving southeast into the lower Mississippi Valley. Page 90 of 211

91 Episode and Date Episode 3: 6/22/2012 Episode 3: 6/23/2012 Episode 3: 6/24/2012 Max Austin 8- Hr Ozone (ppb) and Site Min Austin 8-Hr Ozone (ppb) and Site Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD 64 (DS) 44 (MR) WNW 2.6 ESE 68 (A) 46 (MR) S 1.8 SE 63 (A) 38 (MR) W 1.2 SE Discussion of Selected Large-Scale Weather Features Stationary front extends from a large low pressure system over Hudson Bay down into Northeast Texas. With a surface ridge at the tail end of the front, winds in central TX shift from generally westerly to easterly after noon. The area is on the eastern edge of an upper level high pressure system. Frontal presence over Texas moderates as the large low pressure over Hudson Bay weakens and lifts north. Surface high pressure becomes established over Arkansas, causing southerly flow across central TX. Upper level high moves east to become centered over the Texas Panhandle. Tropical Storm Debby forms in the eastern Gulf of Mexico. Upper level northeasterly flow with high pressure over the Panhandle and low pressure over the western Gulf. T.S. Debby strengthens slightly while moving NNE toward Florida. Surface high pressure remains centered over Arkansas but elongates farther west, creating a more southeasterly flow over central Texas, although winds are out of the west in the morning hours. Page 91 of 211

92 Episode and Date Episode 3: 6/25/2012 Episode 3: 6/26/2012 Episode 3: 6/27/2012 Max Austin 8- Hr Ozone (ppb) and Site Min Austin 8-Hr Ozone (ppb) and Site Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD 75 (SM) 55 (MR) WNW 2.2 NE 87 (NW) 67 (H) W 0.8 ESE 81 (LG) 72 (SM) SE 2.3 SSE Discussion of Selected Large-Scale Weather Features Surface high pressure previously over Arkansas moves to the west to become centered over North Texas. T.S. Debby reaches peak intensity just offshore of the Florida Panhandle and imparts a weak northerly surface flow over central Texas before returning to southeasterly in the late afternoon. Upper level high pressure over the Panhandle persists. Surface high pressure system over north Texas shifts west to the Rockies as a frontal system moves south, allowing development of seabreeze thunderstorms from the Gulf of Mexico to bring precipitation to south Texas. Surface winds during the morning are out of the SW and shift to E or SE later in the day. A surface high pressure system builds in over the southeastern U.S. as T.S. Debby moves away from Florida and into the Atlantic. This sets up a consistent southeasterly flow over central Texas after sunrise. Upper level high pressure reestablishes itself over the Great Plains, sending 100 temperatures as far north as Montana. Page 92 of 211

93 Episode and Date Episode 3: 6/28/2012 Episode 4: 8/9/2012 Episode 4: 8/10/2012 Max Austin 8- Hr Ozone (ppb) and Site Min Austin 8-Hr Ozone (ppb) and Site Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD 61 (A/LG) 53 (H) S 2.4 S 64 (LG) 46 (SM) WSW 2.0 S 81 (LG) 51 (DS) WSW 1.1 WSW Discussion of Selected Large-Scale Weather Features High pressure over southeastern U.S. pushes southward to the Gulf of Mexico while an upper level low pressure develops over central Texas. Deep layer moisture flow from the Gulf of Mexico raises humidity levels across the state. Central Texas sits between two upper level highs: one over the northwestern Gulf and one in the lower Rockies. SW surface winds dominate central Texas as a cool front moves into North Texas. Sea breeze thunderstorm development along the coastal plain. Cool front moves through central Texas and becomes stationary. Ahead of the front, surface winds out of the southeast are caused by high pressure over the northern Gulf of Mexico. Afternoon showers develop along the boundary, which quickly weaken after sunset. Upper air flow from the north is caused by high pressure over the SW U.S. Page 93 of 211

94 Episode and Date Episode 4: 8/11/2012 Episode 4: 8/12/2012 Episode 5: 8/18/2012 Max Austin 8- Hr Ozone (ppb) and Site Min Austin 8-Hr Ozone (ppb) and Site Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD 94 (NW) 61 (SM) NW 0.6 SSE 52 (LG) 38 (H) SSW 3.3 SSE 33 (DS) 24 (H) SSW 1.7 ENE Discussion of Selected Large-Scale Weather Features Cool front weakens and lifts out, with less moisture present than the day before due to the strengthening of an upper level high pressure over West Texas. Surface winds start out northerly but quickly switch to southeasterly around noon. Surface back trajectories suggest that recirculation of polluted air over the monitors along the Interstate 35 corridor is a likely cause of high ozone. A second frontal system makes its way across the Panhandle and into North Texas. Southeasterly flow continues over Central Texas. Upper level high over West Texas remains in place. Upper level NE flow caused by high pressure over NW Mexico. At the surface, the remnants of T.S. Hector were over the northern Gulf of California, sending moist southwesterly flow into Texas and the southeastern U.S. This, combined with a frontal passage through north Texas, causes a wide area of precipitation across the entire southern half of the U.S. Page 94 of 211

95 Episode and Date Episode 5: 8/19/2012 Episode 5: 8/20/2012 Episode 5: 8/21/2012 Episode 5: 8/22/2012 Max Austin 8- Hr Ozone (ppb) and Site Min Austin 8-Hr Ozone (ppb) and Site Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD 44 (MR) 35 (H) NE 1.9 NE 81 (SM) 69 (A) NNW 3.0 NNE 72 (SM) 48 (H) NE 3.1 E 66 (DS) 55 (H) NE 2.1 ENE Discussion of Selected Large-Scale Weather Features Precipitation occurs across central Texas for most of the day. Deep-layer front passes through in the late evening with high pressure building in behind it. Strong northwesterly flow occurs aloft between the high over Mexico and a low over Hudson Bay. NW flow aloft persists. NW flow at the surface dominates before the frontal passage, but gradually shifts to NE throughout the afternoon as the boundary moves slowly south. Passage of the front brings drier, more stable continental air into central Texas. Surface high pressure over west Texas causes NE surface flow across central Texas. Mesoscale convective complex moves from the Panhandle into central Texas by early evening, steered by an upper level ridge centered over northern Mexico, brings gusty winds and a drop in temperatures. Upper level ridge remains in place over northern Mexico, continuing the NW flow aloft over central Texas. Large, east-west elongated surface high pressure becomes established over the middle Mississippi valley. Surface flow NE shifting to E throughout the day. Page 95 of 211

96 Episode and Date Episode 5: 8/23/2012 Episode 6: 9/8/2012 Episode 6: 9/9/2012 Max Austin 8- Hr Ozone (ppb) and Site Min Austin 8-Hr Ozone (ppb) and Site Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD 59 (A/DS) 48 (H) SSE 1.8 E 58 (MR/SM) 49 (H) 83.1 No Data 5.6 N 5.1 N 65 (SM) 50 (H) 87.7 No Data 2.4 NNE 2.8 NE Discussion of Selected Large-Scale Weather Features Surface high over Mississippi valley persists and builds south, causing a SE flow over central Texas, bringing in warm, moist, and cleaner air from the Gulf of Mexico. NW flow continues in the upper levels. Cold front pushes quickly through central TX, causing high temperatures to drop from the upper-90s to the lower-80s. An upper level low is present over extreme northern Mexico, which causes moist SE flow aloft. Surface winds are brisk and northerly following the passage of the front. With the front having cleared Texas, high pressure builds farther east behind it. This causes a shift in surface winds from northerly to NE, but weaker than the previous day. Upper level winds are more stagnant than usual with a large trough over the Great Lakes and Ohio Valley, a high pressure centered over the Gulf of Mexico, and the low pressure over northern Mexico. Page 96 of 211

97 Episode and Date Episode 6: 9/10/2012 Episode 6: 9/11/2012 Episode 6: 9/12/2012 Episode 7: 5/10/2013 Max Austin 8- Hr Ozone (ppb) and Site Min Austin 8-Hr Ozone (ppb) and Site Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD 77 (DS) 57 (H) NE 1.6 ENE 64 (A) 54 (H) SE 2.3 SE 56 (LG) 37 (SM) SSE 2.6 SSE 52 (H) 39 (SM) N 3.0 ENE Discussion of Selected Large-Scale Weather Features Surface high pressure previously over the SE U.S. moves north to be centered over the southern Great Lakes. This reduces its impact on central TX, causing more stagnated conditions, especially in the morning. Upper low over northern Mexico begins to weaken. Upper ridge moves in from NW Mexico to the east as the large trough over the Great Lakes approaches the East Coast. Southeasterly flow returns as high pressure in the eastern U.S. strengthens. Humidity begins to rise as maritime air flows in off the Gulf. In the upper levels, high pressure continues to move east over northern Mexico. Upper level low pressure develops over the lower Mississippi Valley, causing NW flow over central TX. SE flow at the surface continues to bring clean maritime air into central TX. Frontal boundary approaches central TX and brings over an inch of rain to Austin. Weak upper low over the southwest and a weak upper ridge over the southeast cause SW flow aloft over central TX. Surface winds range from NW to NE for most of the day. Page 97 of 211

98 Episode and Date Episode 7: 5/11/2013 Episode 7: 5/12/2013 Episode 7: 5/13/2013 Episode 7: 5/14/2013 Max Austin 8- Hr Ozone (ppb) and Site 64 (MR) Min Austin 8-Hr Ozone (ppb) and Site 52 (LG/SM) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD N 4.4 NNE 60 (MR) 51 (LG) NNE 2.8 NE 70 (A) 66 (SM) S 2.4 SSE 56 (A/LG/H) 51 (SM) SSW 3.7 S Discussion of Selected Large-Scale Weather Features Cold front clears most of Texas and brings N to NE surface flow. Upper level SW flow continues with an elongated trough in the western U.S. Despite the frontal passage, only a modest decrease in temperature and humidity are experienced. N to NE surface flow continues with high pressure over the Midwest, although wind speeds are generally lighter than the previous day. Upper level conditions remain the same except for a deepening low pressure over the northern Great Lakes. Surface high pressure settles into the southeastern U.S. causing very weak southerly flow across central TX. Clear skies and seasonable high temperatures characterize the weather in the Austin area. Upper level flow continues to be out of the SW with an area of low pressure over NW Mexico. Surface high pressure system over the SE U.S. strengthens and moves east, causing stronger S to SW winds across central TX. Page 98 of 211

99 Episode and Date Episode 8: 6/1/2013 Episode 8: 6/2/2013 Episode 8: 6/3/2013 Episode 8: 6/4/2013 Max Austin 8- Hr Ozone (ppb) and Site 39 (LG) Min Austin 8-Hr Ozone (ppb) and Site 28 (NW/DS /MR/SM ) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD S 2.4 S 58 (LG) 49 (DS) NE 2.8 NNE 79 (LG) 59 (DS) E 1.8 SSE 71 (LG) 57 (DS) S 2.7 SSE Discussion of Selected Large-Scale Weather Features Upper level trough in the Midwest extends south to the Panhandle. Across central Texas, a zonal flow is present in the upper levels. At the surface, a cold front pushes south into north Texas. In advance of the front, surface winds are out of the south or southeast. Frontal boundary pushes through central Texas early in the morning leaving a northeasterly flow as high pressure builds in behind it over the Panhandle. Upper level frontal presence not as pronounced as the previous day with zonal flow persisting over central Texas. High pressure over the Panhandle builds east over the lower Mississippi valley. This causes a surface flow reversal from northeasterly to southeasterly over central Texas over the course of the day as the front pushes east. Zonal flow persists in the upper levels. Southern end of the stationary frontal boundary lifts north while surface high pressure over northern Mississippi imparts a steady southerly flow over central Texas. Page 99 of 211

100 Episode and Date Episode 8: 6/5/2013 Episode 8: 6/6/2013 Episode 8: 6/7/2013 Episode 8: 6/8/2013 Max Austin 8- Hr Ozone (ppb) and Site 56 (LG) Min Austin 8-Hr Ozone (ppb) and Site 44 (NW/DS / H) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD S 2.2 SE 64 (LG) 50 (DS) W 0.4 NW 67 (LG) 53 (SM) NNE 1.8 N 71 (LG) 46 (DS) ESE 2.8 S Discussion of Selected Large-Scale Weather Features Mesoscale convective system slides east across the Texas Panhandle. Zonal pattern continues to dominate the upper levels, while at the surface there is persistent SE flow with increased humidity. Mesoscale convective system advances quickly across north Texas during the first half of the day. Frontal boundary is over Central TX causing stagnant winds before noon. After the passage of the front, northerly surface flow becomes established in conjunction with Tropical Storm Andrea in the eastern Gulf of Mexico. Zonal flow occurs in the upper levels. Northerly surface flow continues between the back side of T.S. Andrea and a surface ridge over the northern Great Plains. Passage of the cool front brings a modest drop in humidity and temperatures. Surface winds shift from northerly in the pre-dawn hours to southerly around noon as T.S. Andrea races quickly northeast along the mid-atlantic coast. High pressure builds in behind it and causes S-SE flow over central TX. In the upper levels, a trough begins to form over the lower Mississippi Valley. Page 100 of 211

101 Episode and Date Episode 8: 6/9/2013 Episode 9: 7/1/2013 Episode 9: 7/2/2013 Episode 9: 7/3/2013 Max Austin 8- Hr Ozone (ppb) and Site 54 (LG) Min Austin 8-Hr Ozone (ppb) and Site 33 (DS/SM) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD NW 1.6 E 60 (LG) 47 (MR) NNE 4.8 N 67 (LG) 52 (MR) N 3.0 N 82 (SM) 56 (MR) N 1.9 ENE Discussion of Selected Large-Scale Weather Features Another frontal system approaches Central TX, bringing almost a quarter-inch of precipitation. Front becomes stationary over central TX and causes multiple wind shifts throughout the day. Large deep layer low pressure system over Ohio and Mississippi Valley, together with an upper level ridge over the southern Rockies cause northeasterly flow aloft across central Texas. Stationary front present over the northern Gulf of Mexico and west Texas. High pressure over the northern Great Plains causes N to NE surface flow across central TX. Northerly surface flow continues across central Texas as the low pressure over the Ohio Valley remains stationary. Surface high pressure begins to build in farther south behind the stationary front still over the Gulf of Mexico. The gradual transition back to SE surface flow begins as the surface high behind the front settles in over the lower Mississippi Valley. Northerly surface flow over central Texas begins to weaken after sunrise. Upper level trough still in place over the Great Plains, providing NW flow aloft. Page 101 of 211

102 Episode and Date Episode 9: 7/4/2013 Episode 9: 7/5/2013 Episode 9: 7/6/2013 Episode 9: 7/7/2013 Episode 10: 8/16/2013 Max Austin 8- Hr Ozone (ppb) and Site Min Austin 8-Hr Ozone (ppb) and Site Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD 75 (LG) 55 (MR) SSW 1.2 SE 78 (LG) 58 (MR) SSW 2.1 SE 68 (LG) 54 (MR) SSE 2.7 ESE 46 (LG) 30 (MR) SE 2.2 ESE 60 (LG) 47 (DS) NNW 1.2 NW Discussion of Selected Large-Scale Weather Features Surface high pressure system over the lower Mississippi Valley firmly establishes a general southerly flow over central Texas. Upper level trough still in place over the central U.S. Upper level trough with axis over central Texas begins to weaken. Large high pressure over western Atlantic brings south to southeasterly flow across central Texas. Central TX lies on the eastern edge of an upper level ridge centered over New Mexico. This causes a shift of upper level winds from NW to NE. Surface high over western Atlantic continues to bring S to SE flow over central Texas. Upper level trough lifts north and moves east. Central Texas still on the eastern edge of an upper level ridge. Surface SE flow continues with increased precipitation along the Texas coast and rising humidity. Stationary front is located over the southern part of the state and is accompanied by precipitation and stagnant air flow early in the day. Winds become more northerly later in the day. NW flow dominates the upper levels with a ridge over the SW U.S. and a deepening trough just west of the Mississippi River Valley. Page 102 of 211

103 Episode and Date Episode 10: 8/17/2013 Episode 10: 8/18/2013 Episode 10: 8/19/2013 Episode 11: 8/20/2013 Max Austin 8- Hr Ozone (ppb) and Site Min Austin 8-Hr Ozone (ppb) and Site Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD 70 (LG) 55 (H) NE 3.4 NE 70 (SM) 54 (H) ENE 2.4 NE 66 (A/LG) 51 (H) SSW 1.6 SE 58 (LG) 42 (MR) NNE 3.1 ESE Discussion of Selected Large-Scale Weather Features Surface NE flow continues on the back side of the frontal boundary now over the northern Gulf of Mexico. NW flow aloft continues as trough of low pressure continues to strengthen and move east. Stationary front is still present along the Gulf Coast. Upper level flow becomes more northerly as the ridge over the southwest becomes elongated north-south. Surface wind flow continues to be out of the northeast, although not as strong as the previous day. Upper level trough over the Mississippi Valley lifts out and becomes replaced by a large ridge of high pressure. Surface winds become southerly as the stationary front weakens, allowing high pressure to build into the southeast U.S. A tropical wave combined with the residual frontal boundary causes precipitation along the TX Gulf coast. Surface SE flow continues and an upper level ridge remains in place over much of the country. Page 103 of 211

104 Episode and Date Episode 11: 9/22/2013 Episode 11: 9/23/2013 Episode 11: 9/24/2013 Max Austin 8- Hr Ozone (ppb) and Site Min Austin 8-Hr Ozone (ppb) and Site Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD 60 (LG) 42 (MR) NNE 3.1 NNE 73 (A/LG) 43 (MR) E 1.3 SE 72 (SM) 47 (MR) W 2.8 NNE Discussion of Selected Large-Scale Weather Features Surface high pressure becomes established over west central Texas after the passing of a cold front the day before. Surface low is at the tail end of the front off the Texas coast, causing north to northeasterly winds. Upper level ridge builds in behind the cold front, which becomes stationary over the northern Gulf of Mexico. Pressure gradient steepens aloft while surface winds moderate. Upper level ridge axis crosses central Texas. Surface high pressure forms over the southeastern U.S. while a surface low at the end of the cold front in the western Gulf of Mexico strengthens. These two features cause stagnant wind flow to persist the entire day over central TX. Frontal system passes through central Texas at the tail end of a large low pressure system over the Midwest. Winds in the Austin area are westerly for most of the morning, gradually shifting clockwise to northeasterly in the afternoon. In the upper levels, the high pressure system moves east and a trough moves through the central U.S. This causes NW flow aloft. Page 104 of 211

105 Episode and Date Episode 11: 9/25/2013 Episode 11: 9/26/2013 Episode 11: 9/27/2013 Max Austin 8- Hr Ozone (ppb) and Site Min Austin 8-Hr Ozone (ppb) and Site Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD 89 (LG) 53 (MR) SE 1.5 SSE 69 (LG) 42 (DS/MR) SSW 2.9 SSE 31 (LG) 16 (DS) SSE 3.0 SSE Discussion of Selected Large-Scale Weather Features With the passage of the front, high pressure builds in both aloft and at the surface over central Texas. Northwesterly winds early in the day become calm and then southerly through the afternoon. Surface low in the Gulf of Mexico weakens and moves east. In the upper level, high pressure is centered over central TX. Southerly to southeasterly surface flow becomes established as the surface high pressure system moves northeast into the southeastern U.S. Upper level ridge remains in place over Texas. Humidity is on the increase. Southeasterly flow brings in increased humidity and precipitation from the Gulf of Mexico. Upper level ridge axis begins to move east into Louisiana. Page 105 of 211

106 Episode and Date Episode 12: 5/15/2014 Episode 12: 5/16/2014 Episode 12: 5/17/2014 Episode 12: 5/18/2014 Max Austin 8- Hr Ozone (ppb) and Site 61 (M/LG/ L/SM) 70 (DS/LG) Min Austin 8-Hr Ozone (ppb) and Site Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD 57 (DS) SSW 1.7 SW 55 (MR) SSW 4.5 S 66 (LG) 53 (MR) S 4.1 SSE 58 (LG) 46 (MR) SSE 3.6 SSE Discussion of Selected Large-Scale Weather Features Unseasonably strong frontal boundary pushed through central TX between the 12th and 13th and increased ozone levels from the mid 20s to the upper 50s (ppb). Near-record lows reported. Early morning winds are stagnant as the NW winds behind the front transition to SW later in the day. Strong surface high pressure is centered over central TX. Axis of upper level trough runs down the center of the U.S. and into central TX, providing NW to W upper level flow. As the surface high pressure moves east, southerly winds increase while another frontal boundary becomes stationary over north TX. Steady upper level NW flow is established as the upper trough moves east. Southerly flow continues as surface high pressure builds in the SE U.S. Upper level flow gradually becomes more zonal as the trough weakens and continues moving east. Continued S to SE flow on the backside of surface high over the Southeast. Stationary front remains over north TX. Page 106 of 211

107 In addition to the Table 8-2 weather summaries, AACOG generated five-day HYSPLIT back-trajectories using starting heights of 10 m, 500 m, and 1 km AGL for each day with 8-hour ozone concentrations >=70 ppb. All back-trajectories were initiated at 1600 CST (2200 UTC) at Austin Northwest (CAMS 3) and are shown in Figures 8-1 through These back-trajectories visually summarize the large-scale flow patterns prior to and during high ozone days in the Austin area. In addition, the back-trajectories indicate the potential source regions of background ozone entering Central Texas. Page 107 of 211

108 Figure 7-1. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on May 14, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 108 of 211

109 Figure 7-2. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on May 16, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 109 of 211

110 Figure 7-3. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on May 17, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 110 of 211

111 Figure 7-4. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on May 21, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 111 of 211

112 Figure 7-5. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on June 1, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 112 of 211

113 Figure 7-6. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on June 25, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 113 of 211

114 Figure 7-7. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on June 26, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 114 of 211

115 Figure 7-8. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on June 27, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 115 of 211

116 Figure 7-9. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 10, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 116 of 211

117 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 11, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 117 of 211

118 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 20, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 118 of 211

119 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 21, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 119 of 211

120 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 10, 2012 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 120 of 211

121 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on May 13, 2013 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 121 of 211

122 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on June 3, 2013 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 122 of 211

123 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on June 4, 2013 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 123 of 211

124 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on July 3, 2013 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 124 of 211

125 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on July 4, 2013 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 125 of 211

126 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on July 5, 2013 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 126 of 211

127 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 17, 2013 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 127 of 211

128 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 18, 2013 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 128 of 211

129 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 23, 2013 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 129 of 211

130 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 24, 2013 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 130 of 211

131 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 25, 2013 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 131 of 211

132 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on May 16, 2014 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 132 of 211

133 7.1 Commonly Observed Large-Scale Weather Features The case studies summarized in Table 8-2 reveal common large-scale synoptic weather features that contribute to elevated ozone levels in the Austin area. All of the high ozone episodes during were initiated by the passage of a cold front through Central Texas. Some cold fronts were accompanied by strong gusty winds, precipitation, and the transport of noticeably cooler air into Texas, while other cold fronts primarily represented a transition zone between drier continental air to the north and moister, maritime air to the south. For these latter systems, increased solar radiation and drier air were associated with increases in the daytime maximum temperatures compared to pre-frontal conditions. For some high ozone episodes, the southward movement of a high pressure ridge into Texas behind a cold front was associated with northerly winds throughout the lower troposphere that transported continental air into the Austin area from locations located well north of Texas. Other high ozone episodes were initiated when a surface high pressure ridge over eastern portions of the U.S. expanded southwestward into Texas and the long-range transport of air into the Austin area was from continental regions located well to the northeast or east of Texas. For this latter scenario, the high pressure ridge was usually associated with a cold front that had moved south into or through the eastern U.S. during previous days. This ridge is typically associated with clear skies, warm temperatures, and light wind speeds at the surface. The high pressure acts as a cap to prevent vertical venting of pollutants, keeping them more concentrated closer to the surface. The light winds associated with these high pressure systems also act to inhibit horizontal advection of pollutants, keeping them contained near the urbanized area. High pressure is often, but not always, over Texas at upper levels as well but northerly or northwesterly flow aloft was not uncommon. As the surface ridge of high pressure weakened or moved eastward during the following days, southeasterly and southerly flow throughout the near-surface layer often returned in Central Texas. This flow may have been initially associated with the continued transport of continental air over the Gulf of Mexico and into Texas at one or more levels above the surface, and did not represent an immediate return to relatively clean maritime tropical inflow. With continued flow of relatively clean maritime air from the Gulf of Mexico into Texas, the horizontal dilution of ozone was improved in the Austin area and ozone concentrations decreased. It was also common for a second cold front to move into Central Texas, sometimes initiating an additional multi-day high ozone event. For all high ozone episodes, high ozone concentrations were typically measured at ozone monitors throughout the eastern half of Texas demonstrating the regional nature of high ozone events. The continental air mass transported into Texas likely contained elevated background concentrations of ozone and its precursor compounds associated with both biogenic and anthropogenic emissions. 7.2 Upwind Geographic Source Regions The back-trajectories provided in Figures 8-1 through 8-25 can be used to investigate the synoptic-scale long-range flow patterns that have been associated with the highest ozone concentrations in the Austin area, and visualize the potential source regions of ozone entering the Austin area, with an emphasis on areas located outside of Texas. The back-trajectories are consistent with the clockwise flow around a surface ridge of high pressure that extends south or southwest into Texas. Afternoon mixing heights on high ozone days in Central Texas often reach km AGL. 28 Based on this mixed layer depth, backtrajectories were initiated at 10 m, 500 m, and 1 km AGL at Austin Northwest. With the deepening of the mixed layer during the afternoon hours, ozone from heights up to 2 km above the surface would likely be mixed to the surface and play a role in the ozone concentrations measured in Austin. Out of the 28 Page 133 of 211

134 65 long-range back-trajectories initiated at a starting height of 1 km AGL between 2012 and 2014, only five days had long-range flow from a non-continental region. All but one of these events occurred during May and June. Although the five-day back trajectories for these days originated over or spent most of the time over the Gulf of Mexico, three out of the five days featured a recirculation of air over Central Texas. The relative lack of back-trajectories that originate over maritime regions suggests that the inflow of continental air into Texas at one or more levels above the surface is a necessary condition for high ozone in Austin. Common upwind areas, which identify potential source regions of ozone or its precursors entering Texas, included the Central Plains, Mississippi and Ohio River Valleys, and (less commonly) the Southeastern U.S., although many back trajectories originated as far north as Canada. Individual states most frequently upwind prior to high ozone days in the Austin area were Arkansas, Missouri, Oklahoma, and Louisiana. 8 Meteorological Conditions During Ozone Episodes Currently Used in Selected TCEQ and EPA Modeling Platforms This section reviews meteorological conditions in fall 2006, fall 2011, and June 2012, all of which represent periods of high ozone within the Austin-Round Rock MSA and which photochemical modeling base cases have been developed. CAPCOG has been using the June 2006 base case for photochemical modeling in recent years, and the analyses in this section provide perspective on the meteorological conditions present in these other modeling platforms. 8.1 August October 2006 In addition to the June 2006 base case episode, TCEQ has also developed a base case episode for August 15, 2006 September 15, Similar to the analysis of weather patterns during high ozone episodes that occurred during summarized in Section 8.0 of this report, AACOG also investigated the large-scale atmospheric circulation features during Austin high ozone episodes for August October 2006 using upper air and surface weather maps maintained by the Precipitation Diagnostics Group in the Mesoscale and Microscale Meteorology Division of NCAR. 29 Summaries were generated for each day of the modeling episode with an emphasis on days that had Austin area 8-hour ozone concentrations >= 70 ppb. The results are presented in Table 9-1 and include descriptive discussions of the relevant upper air and surface large-scale weather features in addition to local meteorological observational summaries at Austin Northwest (maximum daily temperature: T, average morning (6a 10a) and afternoon (12p 4p) wind speed: WS, morning and afternoon wind direction: WD). The daily average relative humidity (RH) was obtained using data from TCEQ at Camp Mabry. Also provided in Table 9-2 are the minimum and maximum 8-hour daily maximum ozone concentrations measured by the Austin area monitoring network. The results presented in Table 9-1 are the same analyses that were performed for the representative high ozone episodes during (refer to Table 8-2). These results are provided so that a detailed and direct inter-comparison of conditions on specific August October 2006 days can be made to those that occurred during , if desired. In addition to the Table 9-1 weather summaries, AACOG generated five-day HYSPLIT back-trajectories using starting heights of 10 m, 500 m, and 1 km AGL for each day with 8-hour ozone concentrations >= 70 ppb. All back-trajectories were initiated at 1600 CST at Austin Northwest and are shown in Figures 9-1 through These back-trajectories visually summarize the large-scale flow patterns prior and during 29 Page 134 of 211

135 high ozone days in Austin. In addition, the back-trajectories indicate the potential source regions of background ozone entering Central Texas. The case studies for August October 2006 summarized in Table 9-2 demonstrate that a surface ridge of high pressure extended south or southwest into Texas during high ozone episodes in the Austin area. This ridge was associated with clear skies, warm temperatures, and light wind speeds at the surface. Most of the high ozone events during the August October 2006 episode were initiated with the passage of a cold front into or through Central Texas, which allowed for the transport of continental air from the north as high pressure would become established over the Great Plains or in the northeast, extending SW into Texas. It usually took one or two days for ozone to reach levels above 70 ppb following a frontal passage, mostly due to the effects of precipitation and cloudiness associated with the frontal boundary, thus limiting the solar radiation required to form ozone. With the continued southward movement of these frontal boundaries, high pressure was able to build in the southeastern U.S. With Central TX being on the backside of these high pressure systems, southeasterly flow from the Gulf of Mexico would act to bring ozone concentrations down to near-background levels, where they would stay until another frontal system passed through. There was one high ozone event during the August October 2006 episode that was not immediately preceded by a cold front. October 5 th and 6 th both had 8-hour ozone values of over 70 ppb, but the most recent front to pass through Central TX was on September 28 th. After the relatively quick frontal passage, the northerly winds associated with high pressure over the Great Plains were short-lived and quickly transitioned to SE winds as high pressure moved over the Ozarks. The high pressure system continued moving east into the mid-atlantic and extended SW to central TX. As another frontal system passed north of Texas, it weakened the western edge of the high, causing more stagnant conditions and allowing ozone to form and accumulate in the absence of any horizontal advection. These days are not included in TCEQ s fall 2006 model. Table rows are color-coded according to the maximum ozone concentration: Blue: ppb; Green: ppb; and Yellow: 75 ppb or higher. Site abbreviations are as follows: NW: Austin Northwest (CAMS 3); A: Audubon (CAMS 38); P: Pflugerville (CAMS 613); DS: Dripping Springs (CAMS 614); RR: Round Rock (CAMS 674); SM: San Marcos (CAMS 675); and MR: McKinney Roughs (CAMS 684). Page 135 of 211

136 Table 8-1. Daily local meteorological measurements at Austin Northwest and large-scale weather features for August October 2006 Date 8/29/2006 8/30/2006 Max Austin 8-Hr Ozone (ppb) and Site 46 (RR / MR) 70 (NW/ A) 8/31/ (A) 9/1/ (NW) Min Austin 8-Hr Ozone (ppb) and Site 31 (DS) 63 (SM) 63 (SM) 64 (SM) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD NNW 2.1 NNE 89.7 No Data 3.4 NE 1.8 NNE WNW 0.8 E SW 1.3 N Discussion of Selected Large-Scale Weather Features Heavy rainfall moves across Texas with the passage of a cool front in conjunction with an upper level low off the Gulf coast of Mexico that provides deep layer moisture. North to northeasterly surface winds are caused by a large surface high pressure centered over the Great Lakes but extending SW into the TX Panhandle. Winds at the surface are slightly more stagnant than the previous day. Upper low in the NW Gulf of Mexico and an upper level ridge over northern Mexico bring NW winds aloft. Upper level trough extends from the Great Lakes down the Mississippi valley, ushering in continental air from the Rockies. At the surface, steering currents are very light as central TX sits between a large surface high over Hudson Bay, a smaller high pressure area over the Rockies, and T.S. Ernesto off the east coast. Stagnant conditions continue across central TX. Upper level ridge builds over northern Mexico and establishes a SE flow aloft. Long range back trajectories show transport from the Great Lakes. Surface flow is generally out of the SW to SE. Page 136 of 211

137 Date 9/2/2006 Max Austin 8-Hr Ozone (ppb) and Site 78 (NW) 9/3/ (P) 9/4/2006 9/5/ (NW) 49 (RR) Min Austin 8-Hr Ozone (ppb) and Site 58 (DS) 54 (DS) 51 (DS) 30 (SM) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD SW 0.9 NNE WSW 1.1 NE NE 1.5 NNE NNE 2.3 NNE Discussion of Selected Large-Scale Weather Features Cold front advances through north TX with plenty of moisture aided by T.S. John located over the southern Baja Peninsula. Upper level ridge is located over much of Texas. Surface ridge in the western Gulf continues to bring southerly flow across central TX which switches to northerly in the afternoon. Cold front stalls over west central Texas. Widespread showers are behind the boundary with southerly flow ahead of the front. Winds switch to northerly after noon which may have caused recirculation of ozone and precursors over the Austin area. Upper level ridge is still located over northern Mexico. Surface winds are northerly throughout most of the day, indicating the stationary front had passed through. Upper level ridge continues over northern Mexico m back trajectories indicate transport from the Ohio Valley. Upper level ridge shifts southward into central Mexico as the cold front finally pushes through the area. High pressure builds in behind the front, centered over the Great Plains. Moisture associated with remnants of T.S. John brings widespread showers to central TX. Page 137 of 211

138 Date Max Austin 8-Hr Ozone (ppb) and Site 9/6/ (P) 9/7/ (NW) 9/8/ (P) 9/9/ (P) 9/10/ (RR) Min Austin 8-Hr Ozone (ppb) and Site 59 (DS) 71 (SM) 55 (DS / SM) 42 (DS) 34 (DS) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD N 3.4 NNE NNW 1.7 NNE S 1.7 SE ESE 2.2 ESE S 0.4 SE Discussion of Selected Large-Scale Weather Features Northerly surface winds continue as N-S elongated high pressure remains over Great Plains. Upper level trough covers much of the central U.S. but is elongated NE to SW. Five day back trajectories in the upper levels originate in the Great Lakes. Strong downward vertical motion is observed in back trajectories. Surface high pressure over the central U.S. is now centered over the Mid Atlantic, but extends to the Rockies. Back trajectories continue to show transport from the Great Lakes and Ohio Valley at all levels. Moist southerly flow returns to central TX as the surface high pressure moves into the SE U.S. In the upper levels, a more zonal flow becomes established as the cold front weakens. Similar pattern, but with a more easterly component to the surface winds. With more moisture present, showers form at the tail end of the remnant frontal boundary along the TX Gulf coast. Cloudy conditions keep temperatures in the low 80 s and preclude ozone formation. Another cold front approaches north TX. Continued flow of moisture from the Gulf triggers afternoon showers across the state. Zonal upper air pattern continues. Page 138 of 211

139 Date Max Austin 8-Hr Ozone (ppb) and Site 9/11/ (A) 9/12/2006 9/13/ (RR) 66 (NW) 9/14/ (P) Min Austin 8-Hr Ozone (ppb) and Site 12 (SM) 24 (SM) 54 (DS) 63 (SM) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD SSW 1.2 S WNW 1.4 ESE NNE 3.0 NNE NNE 1.6 SSE Discussion of Selected Large-Scale Weather Features Cold front moves through the Panhandle and into north TX. Scattered showers continue during the afternoon with almost a half inch of rain reported in Austin. Surface winds shift from southerly to northerly as the front moves slowly through central TX. Upper level ridge over NE Mexico and upper level low over Arizona help transport tropical moisture aloft. Although ample moisture is present, Austin only receives a trace of rain. Cold front clears central TX. Winds are northerly and stronger than the previous day due to the location of the surface high over north TX. Zonal upper air pattern continues with the base of the trough associated with the cold front only extends to Arkansas. Northerly winds dominate the morning hours as the high pressure that builds in behind the cold front is positioned north of Austin. Southeasterly flow returns after noon as the high migrates to the northeast to become centered over the Great Lakes. Page 139 of 211

140 Date Max Austin 8-Hr Ozone (ppb) and Site 9/15/ (P) 9/16/ (P) 9/17/ (A) 9/18/ (P) 9/19/ (DS) Min Austin 8-Hr Ozone (ppb) and Site 26 (DS / SM) 23 (DS) 22 (SM) 38 (SM) 55 (SM) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD S 3.1 SSE S 3.8 S S 3.4 S NNW 3.8 NNW NNE 2.7 ENE Discussion of Selected Large-Scale Weather Features High pressure continues over the lower Great Lakes. Surface low pressure deepens along a cold front located in the NW U.S, intensifying the southerly flow across central TX. Humidity is on the increase and scattered showers form along the coast. Brisk southerly winds continue between low pressure east of the Rockies and high pressure over the Great Lakes. Cold front extending from the low pressure approaches Texas. Upper level ridge of high pressure develops over the Gulf. Southerly winds continue, providing moisture for the cold front now located over north TX. Rainfall almost a quarter inch in Austin. Upper level ridge intensifies over the northern Gulf. Cold front clears central TX with brisk northerly winds and drier air. Rainfall continues through the morning with over 0.8 inches recorded. Northerly surface winds become more easterly as the high pressure behind the cold front moves east to be centered over the southern Great Plains. Overall, wind speeds are more relaxed than the previous day. Axis of the upper level trough associated with the cold front passes central TX. Page 140 of 211

141 Date 9/20/2006 Max Austin 8-Hr Ozone (ppb) and Site 74 (A/P) 9/21/ (A) 9/22/ (A) 9/23/ (A) 9/24/ (P / MR) Min Austin 8-Hr Ozone (ppb) and Site 57 (SM) 36 (SM) 27 (DS) 30 (MR) 34 (DS) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD SE 2.9 SSE S 4.5 SSW S 4.5 S SSW 2.7 SSW NNE 4.6 N Discussion of Selected Large-Scale Weather Features As the surface high pressure continues moving east into the lower Mississippi valley, winds return to the SE. Morning wind speeds are light as the flow shifts directions from NW to SE. Upper level ridge is located over south TX. Area of low pressure develops over the Rockies and moves ENE across the Great Plains while strengthening in the upper and lower levels. This causes southerly winds to increase as the low passes to the north. Cold front associated with the low crosses the TX Panhandle. Area of low pressure moves into the upper Mississippi valley. Brisk southerly flow continues in central TX in advance of the associated cold front. Upper level low pressure becomes absorbed in the larger upper level trough of low pressure, which covers much of the central U.S. Cold front continues its advance through central TX. Northerly wind shift occurs just after noon along with a drop in temperatures. Over an inch of rain reported at Camp Mabry with the passage of the front. With complete passage of the cold front, drier air sets in and strong northerly winds continue. This is one of the stronger cold fronts of the season with afternoon high temperatures in the mid-70s. Page 141 of 211

142 Date 9/25/2006 9/26/2006 9/27/2006 9/28/2006 Max Austin 8-Hr Ozone (ppb) and Site 59 (NW) 72 (NW) 70 (NW) 56 (MR) Min Austin 8-Hr Ozone (ppb) and Site 47 (DS) 57 (SM) 55 (DS) 44 (RR) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD NNW 1.7 NNE W 1.2 ENE SSW 4.2 SSW NNE 5.0 NE Discussion of Selected Large-Scale Weather Features Northerly surface winds persist but not as strong as the previous two days. Upper level pattern over central TX becomes more zonal as the upper trough moves east. High pressure in place over central TX provides a stagnated air mass with moderating temperatures, variable winds, and continued low humidity. 24-hour back trajectories indicate transport from DFW occurred. Surface high pressure shifts SE to be centered over the northern Gulf, reestablishing a southerly wind flow around the back side. Back trajectories indicate a recirculation of air over the Austin area, which likely contributed to high ozone levels. Another cold front moves through the TX Panhandle. Cold front moves through central TX with winds shifting out of the north during the morning hours. Base of the upper level trough associated with the frontal system extends into the Gulf of Mexico. A trace of rain is reported at Camp Mabry. Page 142 of 211

143 Date Max Austin 8-Hr Ozone (ppb) and Site 9/29/ (A) 9/30/ (A) 10/1/ (PF) 10/2/ (P) 10/3/ (A / P) Min Austin 8-Hr Ozone (ppb) and Site 48 (SM) 34 (DS) 36 (DS) 38 (SM) 39 (SM) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD E 2.8 S SSW 3.3 S S 3.8 S S 3.3 S S 2.8 SSE Discussion of Selected Large-Scale Weather Features Surface winds during the morning hours are northeasterly, shifting to southerly after noon. Northwesterly upper air flow is caused by the upper level trough moving east across the U.S., with the axis along and east of the Mississippi River. Southerly flow continues throughout Texas as high pressure builds in over the northern Gulf. Another frontal system approaches the TX Panhandle. Upper level high pressure located over the southwest U.S. moves east. Upper level ridge of high pressure is now centered over the northwest Gulf. Frontal system previously in the Panhandle has been weakening and is no longer discernible on surface plots west of the Mississippi River. Southerly flow continues around the back side of the surface high over the northern Gulf. Surface high pressure now located over the SE U.S. provides a more SE component to surface flow. Small upper level low pressure develops over central TX and moves to the northwest. Otherwise, in the upper levels, high pressure remains in place along the Gulf coast. Page 143 of 211

144 Date Max Austin 8-Hr Ozone (ppb) and Site 10/4/ (A) 10/5/ (A) 10/6/ (MR) Min Austin 8-Hr Ozone (ppb) and Site 32 (SM) 50 (SM) 63 (SM) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD SSE 0.9 SE WNW 1.0 E NNW 2.7 NE Discussion of Selected Large-Scale Weather Features Surface ridge still located over the SE U.S. weakens, causing a more stagnant air flow over central TX. A cold front approaches north TX. Large upper level high pressure covers all of Texas and the SE U.S. Stagnant air flow continues over central TX with multiple wind shifts throughout the day. 5-day HYSPLIT back trajectories indicate very little air flow in the lower and upper levels. Aloft, a broad area of high pressure is still located over the northern Gulf coast. At the surface, high pressure located over the SE U.S. imparts a weak SE flow, but as a cold front approaches central TX the winds become more variable. Cold front moves through central TX around noon and brings stronger northerly winds and dry continental air. In the upper levels, a ridge of high pressure centered over central TX prevents significant rain from developing. 5-day HYSPLIT back trajectories reflect the different upper and lower level patterns with 1000 m trajectories originating from the SW U.S. and rotating clockwise into central TX. Mid level (500 m) trajectories originate from the northern Great Lakes, and surface trajectories are relatively stagnant and originating from the Houston area. Page 144 of 211

145 Date Max Austin 8-Hr Ozone (ppb) and Site 10/7/ (A) 10/8/ (A) 10/9/ (A) 10/10/ (P) Min Austin 8-Hr Ozone (ppb) and Site 60 (DS) 59 (SM) 48 (SM) 22 (SM) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD ENE 1.6 ENE SSE 1.8 ENE E 1.7 SE WNW 1.0 WNW Discussion of Selected Large-Scale Weather Features Surface winds are generally out of the E or NE due to a NE-SW elongated surface high pressure area building in behind the cold front from New England down to Texas. Upper level high pressure previously over central TX shifts east providing a southerly upper air flow. Surface high pressure over the eastern U.S. weakens and provides a more stagnant air flow over central TX. Surface winds are westerly in the morning, switching to easterly in the afternoon. Upper level high pressure also weakens over central TX. Cold front is over the Texas Panhandle. Stagnant and variable winds early in the morning become steady SE winds after noon due to high pressure over the SE U.S. Upper level high pressure once over central TX continues to weaken and shift east. Cold front makes it to central TX and stalls out. Abundant moisture associated with its passage due to an upper level low located over the extreme SW U.S. drawing in moisture from a developing tropical cyclone in the Eastern Pacific. Over an inch of rain reported at Camp Mabry. Page 145 of 211

146 Figure 8-1. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 30, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 146 of 211

147 Figure 8-2. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 31, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 147 of 211

148 Figure 8-3. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 1, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 148 of 211

149 Figure 8-4. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 2, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 149 of 211

150 Figure 8-5. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 3, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 150 of 211

151 Figure 8-6. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 4, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 151 of 211

152 Figure 8-7. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 6, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 152 of 211

153 Figure 8-8. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 7, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 153 of 211

154 Figure 8-9. Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 8, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 154 of 211

155 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 14, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 155 of 211

156 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 20, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 156 of 211

157 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 26, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 157 of 211

158 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 27, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 158 of 211

159 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on October 5, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 159 of 211

160 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on October 6, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 160 of 211

161 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on October 8, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 161 of 211

162 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on October 9, 2006 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 162 of 211

163 8.2 August October 2011 The EPA has developed a photochemical modeling platform using a 2011 ozone season base case, from May 1, 2011 September 30, AACOG also investigated the large-scale atmospheric circulation features during Austin area high ozone episodes for August October 2011 using upper air and surface weather maps maintained by the Precipitation Diagnostics Group in the Mesoscale and Microscale Meteorology Division of NCAR. 30 Summaries were generated for each day of the modeling episode with an emphasis on days that had Austin 8-hour ozone concentrations >= 70 ppb. The results are presented in Table 9-2 and include descriptive discussions of the relevant upper air and surface large-scale weather features in addition to local meteorological observational summaries at Austin Northwest (maximum daily temperature: T, average morning (6a 10a) and afternoon (12p 4p) wind speed: WS, morning and afternoon wind direction: WD). The daily average relative humidity (RH) was obtained using data from TCEQ at Camp Mabry. Also provided in Table 9-2 are the minimum and maximum 8-hour daily maximum ozone concentrations measured by the Austin area monitoring network. The results presented in Table 9-2 are the same analyses that were performed for the representative high ozone episodes during (refer to Table 8-2). These results are provided so that a detailed and direct inter-comparison of conditions on specific August October 2011 days can be made to those that occurred during , if desired. In addition to the Table 9-2 weather summaries, five-day HYSPLIT back-trajectories using starting heights of 10 m, 500 m, and 1 km AGL were generated for each day with 8-hour ozone concentrations >= 70 ppb. All back-trajectories were initiated at 1600 CST at Austin Northwest and are shown in Figures 9-18 through These back-trajectories visually summarize the large-scale flow patterns prior and during high ozone days in Austin. In addition, the back-trajectories indicate the potential source regions of background ozone entering Central Texas. In contrast to the August October 2006 episode, the August October 2011 episode did not feature as straightforward a weather pattern. In fact, the meteorological conditions present during the first half of this episode were some of the most extreme Central TX has ever seen. The summer of 2011 was characterized by all-time record drought and heat throughout most of Texas. 31 Many days in Central Texas had clear or mostly clear skies, light near-surface wind speeds, and exceptionally warm temperatures; however, high ozone concentrations during mid-june to mid-august did not occur, and peak 8-hour ozone concentrations were typically ppb. A persistent pattern of southeasterly flow off of the Gulf of Mexico was caused by high pressure aloft over much of the southern U.S. in conjunction with surface high pressure over the southeastern U.S. This setup brought cleaner maritime air to the Central TX region during most of the time period leading up to the episode. However, during late August, a change in the predominant weather pattern allowed the large-scale transport of continental air southward into Texas beginning on August 26th (refer to Table 9-2), initiating several days with 8-hour ozone concentrations of at least 70 ppb in Austin. Not every high ozone event during the 2011 episode was immediately preceded by a cold front. The first half of the episode only featured one frontal passage that completely cleared Central TX, and it was associated with the landfall of Tropical Storm Lee in the northern Gulf of Mexico. Although there was only one frontal passage, there were thirteen days with 8-hour ozone of at least 70 ppb. Many of these early-episode high ozone days were associated with wind shifts near the surface from the morning to the afternoon hours, causing a recirculation of pollutants, including ozone, over Central TX. This Page 163 of 211

164 phenomenon is similar to that seen in Houston caused by the sea breeze effect and has been known to occur in the San Antonio area as well on high ozone days Beginning on September 3 rd, as Tropical Storm Lee was approaching the Louisiana coast, northerly surface winds transported continental air into Central TX and contributed to elevated ozone. With the exception of the 5 th, every day between the 4 th and the 13 th saw 8-hour ozone over 70 ppb. Due to the steep pressure gradient between Tropical Storm Lee and high pressure behind the front, winds were quite strong and likely prevented accumulation of high levels of ozone on the 5 th. As winds relaxed but northerly flow persisted, ozone levels rose again to 70 ppb until high pressure became reestablished over the southeastern U.S. The movement of Tropical Storm Lee through the Gulf also served to weaken the upper level ridge that had been planted over the southern U.S. and northern Mexico throughout the summer. As Tropical Storm Lee continued to move off to the northeast, high pressure began to build over Texas at the surface, while in the upper levels, a low pressure system over the Mississippi Valley brought NW winds aloft. By the 13 th, the influence of low pressure aloft diminished as high pressure once again established itself over Texas and northern Mexico. At the surface, high pressure moved into the southeastern U.S., setting up a southeasterly flow off of the Gulf that had been present during July and most of August. This brought ozone concentrations down to their lowest levels of the entire episode. A cold front came close to clearing central TX on the 15 th, but due to the blocking high aloft it stalled out and retreated northward. This created an opportunity for the next cold front to finally clear the area on the 19 th. The wind shift associated with this front was short-lived, lasting only for a day. However, the next few days had more stagnant and variable winds, allowing for elevated ozone through the 24 th. The days where winds shifted 180 degrees were the days with higher ozone. Another front moved through central TX on the evening of the 22 nd. The wind shift associated with this front was slightly longer-lived than the previous front, but the timing of the passage was such that the wind shift occurred in the middle of the day and allowed for slightly higher ozone concentrations than the previous day. Once high pressure became re-established over the southeast, ozone levels moderated to the lower 50s ppb until the final frontal passage of this episode on the 30 th. Wind shifts and stagnant air flow were likely responsible for the elevated ozone reported on October 2 nd and 3 rd. Table rows are color-coded according to the maximum ozone concentration: Blue: ppb; Green: ppb; and Yellow: 75 ppb or higher. Site abbreviations are as follows: NW: Austin Northwest (CAMS 3); A: Audubon (CAMS 38); 32 Cowling, Ellis B., et al., Cari Furiness, Basil Dimitriades, Southern Oxidants Study Office of the Director at North Carolina State University, and David Parrish, Earth System Research Laboratory, National Oceanic and Atmospheric Administration, 31 October 2006 [8 November revision]. Preliminary Findings from the Second Texas Air Quality Study (TexAQS II). A Report to the Texas Commission on Environmental Quality by the TexAQS II Rapid Science Synthesis Team TCEQ Contract Number p. 21. Available online: Accessed 08/06/ Alamo Area Council of Governments, Conceptual Model Ozone Analysis of the San Antonio Region Updates through Year 2010 Available online: Accessed 08/06/2015. Page 164 of 211

165 DS: Dripping Springs (CAMS 614); MR: McKinney Roughs (CAMS 684); MR: McKinney Roughs (CAMS 690); SM: San Marcos (CAMS 675 or CAMS 1675); and H: McKinney Roughs (CAMS 6602). Page 165 of 211

166 Table 8-2. Daily local meteorological measurements at Austin Northwest and large-scale weather features for August October 2011 Episode or Date 8/25/2011 Max Austin 8-Hr Ozone (ppb) and Site 61 (LG / DS) 8/26/ (LG) Min Austin 8-Hr Ozone (ppb) and Site 46 (SM) 56 (SM) 8/27/ (DS) 68 (A) 8/28/ (DS) 76 (A) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p- 4p WS (m/s) 12p- 4p WD SSE 2.6 SE W 1.8 SSE WNW 1.1 NE NW 3.0 NE Discussion of Selected Large-Scale Weather Features A summer of record heat and drought continues as a subtropical ridge centered over New Mexico and NW TX continues to dominate; however, a short wave moving S from N TX is associated with a line of isolated southward moving thunderstorms through Central and SE TX. Dry air at low levels enters TX from the N associated with the clockwise circulation around the subtropical ridge of high pressure that extends S into TX from the Central Plains. High pressure at the surface and upper levels is likely associated with strong subsidence but strong thermals likely mixes downward midtropospheric (drier) air to the surface. Max temperatures across TX are > 100F with clear skies and light wind speeds. A weak surface trough of low pressure north of Central TX supports SW daytime winds. The subtropical ridge of high pressure remains over TX and the combination of atmospheric subsidence, clear skies, and transport of dry continental air into TX from the N results in record high temperatures well in excess of 100F across much of TX. Daytime winds remain low. The record dry and hot conditions across TX continue and nighttime temperatures in Central TX only dip into the upper 80s with daytime temperatures >110 F. Page 166 of 211

167 Episode or Date Max Austin 8-Hr Ozone (ppb) and Site 8/29/ (LG) 8/30/ (LG) 8/31/ (LG) 9/1/ (MR) 9/2/ (DS) 9/3/ (NW) Min Austin 8-Hr Ozone (ppb) and Site 61 (SM) 54 (SM) 49 (SM) 37 (SM) 45 (SM) 57 (MR) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p- 4p WS (m/s) 12p- 4p WD SSW 2.4 SSE S 2.7 SSE S 3.0 SSE SSW 2.1 ESE NE 3.3 ENE NNE 5.0 NNE Discussion of Selected Large-Scale Weather Features Another day of record heat across TX as high pressure over the state finally begins to weaken. Gradual transport of air inland from the Gulf of Mexico begins to increase low-level moisture in coastal TX. A weak low pressure trough dissipates over N TX causing some clouds and showers. S flow results in increasing low-level moisture in Central TX over the next several days as the upper-level high pressure ridge moves SW to NW Mexico. Steady SE flow from a large surface high pressure over the western Atlantic and eastern U.S continues to bring increased low level moisture to Central TX. Upper level ridge over N Mexico begins to elongate east-west. Humidity in central Texas is on the rise as a tropical wave approaches the Texas coast in conjunction with continued SE flow off of the Gulf. Western side of the upper level high weakens, leaving it centered near the TX/OK border. Gulf of Mexico disturbance becomes T.S. Lee. Pressure gradient between it and the large surface high pressure steepens, causing stronger winds over much of Texas and a shift in wind direction from SE to NE. SE winds early in the day give way to NE winds once again as T.S. Lee approaches the central LA coast. Winds increase throughout the afternoon. Central TX is on the eastern edge of an upper level high pressure system that brings in continental air aloft. Page 167 of 211

168 Episode or Date 9/4/2011 Max Austin 8-Hr Ozone (ppb) and Site 70 (NW / DS) 9/5/ (H) 9/6/2011 9/7/ (SM) 86 (DS / SM) Min Austin 8-Hr Ozone (ppb) and Site 62 (SM) 48 (SM) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p- 4p WS (m/s) 12p- 4p WD NNW 7.8 N N 5.6 N 59 (A) NNE 1.9 NE 74 (H) N 2.0 NNE Discussion of Selected Large-Scale Weather Features N to NE winds dominate at all levels. T.S. Lee moves onshore in central LA. Humidity continues to decline as drier air from the north is pulled into central TX with the influence of the upper level high over northern Mexico and T.S. Lee to the east. Sunny conditions prevail as the cloud shield of T.S. Lee ends just to the west of Houston. A first strong cold front enters TX after a record dry and hot summer. The cold front is associated with gusty winds and near-record low temperatures. At upper levels, high pressure remains centered over N Mexico while a trough over E TX moves eastward. Skies are clear over TX although smoke from wildfires (including Bastrop State Park) impacts visibility over portions of Central TX. With the transport of dry continental air into TX from the north, relative humidity is extremely low and morning low temperatures are in the 50s with near-normal highs in the 80s. A high pressure ridge at the surface extends from the Great Lakes SW into TX. Smoke from wildfires continues to impact visibility in Central TX. Stagnant atmospheric conditions and light NW winds continue in association with the surface ridge of high pressure. Nighttime temperature reach nearrecord lows in Central TX under continued cloudless skies. Page 168 of 211

169 Episode or Date 9/8/2011 9/9/2011 9/10/2011 9/11/2011 Max Austin 8-Hr Ozone (ppb) and Site 71 (H / MR) 78 (SM) 75 (SM) 82 (SM) Min Austin 8-Hr Ozone (ppb) and Site Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p- 4p WS (m/s) 12p- 4p WD 66 (A) N 3.4 N 65 (A) W 1.3 N 57 (MR) SSW 2.3 NNE 68 (A) SW 0.5 ESE Discussion of Selected Large-Scale Weather Features Weak high pressure at the surface is associated with continued light N flow into TX. High level cloudiness is associated with confluent flow aloft between an Ohio River Valley upper low and the high pressure ridge centered over N Mexico. Large diurnal temperature ranges continue with continued dry N flow of air into TX. Tropical storm Nate moving west over southern Gulf. Stagnant atmospheric conditions over TX persist and are associated with clear skies, warming temperatures, and no precipitation. The large-scale counterclockwise circulation around the upper level low over the Ohio River Valley continues to enhance the transport of dry continental air into TX from the N and NE. As the upper level ridge centered over N Mexico expands northward, upper level winds over TX become W compared to N during previous days. A weak stationary front lies along the Red River Valley while Tropical Storm Nate located well south of TX limits the return flow of moisture into TX from the Gulf of Mexico. Central TX remains dry and hot. Conditions over Central TX change little as the upper level ridge remains over TX with mostly clear skies. Nighttime temperatures are near normal with high temperatures near 100 F. Daytime winds are light from the SW. Page 169 of 211

170 Episode or Date Max Austin 8-Hr Ozone (ppb) and Site Min Austin 8-Hr Ozone (ppb) and Site 9/12/ (H) 68 (A) 9/13/ (H) 9/14/ (LG) 9/15/ (A) 9/16/ (LG/H) 59 (SM) 58 (MR) 58 (MR) 45 (DS) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p- 4p WS (m/s) 12p- 4p WD SSW 2.5 S SW 3.0 S SSW 3.0 SSE NNW 2.1 SE SE 3.2 S Discussion of Selected Large-Scale Weather Features High pressure over TX strengthens as an upper level trough digs eastward over the E US and a cold front moves S from the N US. At the surface, a high pressure ridge extending from the SE US into TX is associated with SW winds over much of the state. Conditions remain clear, dry, and very hot with maximum temperatures near 100 F. Surface winds in Central TX remain SW as a cold front moves S from N TX. At upper levels, high pressure remains over TX centered over N Mexico while a low pressure trough is off the E US coast. The weak cold front moves slowly S through TX reaching N Central TX and is associated with some clouds and showers The weak cold front associated with some clouds and showers becomes stationary across Central TX. Weakening high pressure at upper levels and southeasterly low-level flow from the Gulf of Mexico is associated with increasing relative humidity and high temperatures in the mid-90s. Humidity is on the rise as SE flow off the Gulf continues. Stationary front over central TX causes showers to develop and move east across the area as a zonal upper level pattern emerges with the weakening of the high. Page 170 of 211

171 Episode or Date Max Austin 8-Hr Ozone (ppb) and Site 9/17/ (H) 9/18/ (LG) 9/19/2011 9/20/ (H/MR) 79 (NW) 9/21/ (LG) 9/22/2011 9/23/ (H/MR) 64 (SM) Min Austin 8-Hr Ozone (ppb) and Site 37 (NW) 44 (MR / DS) 59 (DS / A) 71 (LG) 61 (SM) 64 (LG) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p- 4p WS (m/s) 12p- 4p WD SSE 2.2 SSW S 2.6 S NNW 2.7 NNE W 0.5 ENE SSW 1.6 SE WSW 2.1 NNW 55 (A) NNE 2.0 N Discussion of Selected Large-Scale Weather Features Zonal upper level flow persists. Showers and thunderstorms continue across central TX as the frontal boundary lingers. Surface high pressure over SE U.S. continues to bring in moisture from the Gulf. An upper level trough over TX and increasing clouds and showers are associated with an approaching cold front during the evening and nighttime hours. Surface winds turn northerly as a cold front slowly moves through Central Texas. At upper levels, mostly zonal flow dominates over TX. A weak ridge of high pressure extends SW into TX from the Great Lakes. Continued northerly winds transport drier continental air into TX from the northeast. The cold front is now located along coastal TX. The cold front along coastal TX dissipates and winds turn southerly in Central Texas ahead of a second approaching cold front. A second cold front moves slowly through Central TX. Northerly surface winds are associated with the clockwise circulation around a center of high pressure that extends south into TX from the Central Plains. A deep trough of low pressure extends southward into Central TX at upper levels as the cold front continues to move southward offshore of coastal TX. Northerly winds bring drier and colder air into TX. Page 171 of 211

172 Episode or Date 9/24/2011 Max Austin 8-Hr Ozone (ppb) and Site 79 (H/SM) 9/25/ (H) 9/26/ (LG) 9/27/ (DS) 9/28/ (H) Min Austin 8-Hr Ozone (ppb) and Site Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p- 4p WS (m/s) 12p- 4p WD 68 (A) SE 1.4 S 59 (MR) 48 (SM) 45 (SM / H) 55 (NW / A / SM) SSW 4.6 SSW SSE 4.0 S SSW 2.3 SSE SSW 0.8 SW Discussion of Selected Large-Scale Weather Features At upper levels, winds over TX are generally northerly associated with the counterclockwise flow around an upper level trough of low pressure. Southerly winds in Central TX area associated with the clockwise flow of air around the southwestward portion of the surface ridge of high pressure that extends SW into TX from the Ohio River Valley. Relative humidity remains low as daytime temperatures increase. Daytime temperatures exceed 100 F as a second upper level trough moves into NW TX. At the surface, high pressure weakens over TX in advance of a second cold front moving southward from N TX. Winds at the surface in Central Texas are southerly to southsouthwesterly. Cold front stalls across central TX. Surface winds range from SE to SW, bringing increased low-level moisture. Light showers develop along the boundary during the evening. With humidity on the rise, clouds and showers become more numerous across central TX as the front remains stationary over the area. Upper level high pressure begins to advance eastward from the Pacific Ocean. Weakening stationary front continues to trigger scattered showers across SW to central TX. Surface winds shift slightly to range from NW counterclockwise to S and act to lower the humidity. Page 172 of 211

173 Episode or Date Max Austin 8-Hr Ozone (ppb) and Site 9/29/ (H) 9/30/ (DS) 10/1/ (DS) 10/2/ (DS) 10/3/ (DS) Min Austin 8-Hr Ozone (ppb) and Site 41 (MR) 54 (LG) 55 (NW / A / LG) 64 (LG) 65 (MR) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p- 4p WS (m/s) 12p- 4p WD SSW 1.1 SSW N 5.1 NNE NE 1.7 NE WNW 1.4 ESE SSE 1.5 SE Discussion of Selected Large-Scale Weather Features New cold front approaches N TX as the existing front continues to weaken. Surface high pressure in the Gulf continues southerly flow into central TX. Widespread showers develop in the afternoon as the cold front approaches. With upper level high building in the SW U.S. and the passage of a strong cold front, continental air is transported into central TX at all levels. Cold front brings highs in the mid-80s and lows in the mid-60s. Generally northerly winds persist throughout the day. Large surface high pressure builds in behind the cold front and extends from the Great Lakes into the Gulf. Very dry conditions and seasonable temperatures accompany the front. Strong NW winds aloft are caused by upper level high pressure over the southern Rockies and a large trough of low pressure over the Mid Atlantic. As the cold front continues advancing south and surface high pressure builds in behind it, surface winds shift from northwesterly to southeasterly, causing recirculation of emissions over Austin and leading to high ozone. Upper level ridge over west TX continues to drive continental air from the north into central TX. Surface winds now out of the SE as high pressure persists over the SE U.S. Page 173 of 211

174 Episode or Date Max Austin 8-Hr Ozone (ppb) and Site 10/4/ (DS) Min Austin 8-Hr Ozone (ppb) and Site 57 (MR / LG) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p- 4p WS (m/s) 12p- 4p WD ESE 2.5 SE Discussion of Selected Large-Scale Weather Features Large low pressure system moves into Atlantic. SE surface winds continue, bringing clean maritime air from the Gulf. This pattern continues over the next week. Page 174 of 211

175 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 26, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 175 of 211

176 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 27, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 176 of 211

177 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 28, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 177 of 211

178 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on August 29, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 178 of 211

179 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 4, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 179 of 211

180 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 6, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 180 of 211

181 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 7, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 181 of 211

182 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 8, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 182 of 211

183 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 9, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 183 of 211

184 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 10, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 184 of 211

185 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 11, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 185 of 211

186 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 12, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 186 of 211

187 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 13, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 187 of 211

188 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 20, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 188 of 211

189 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 22, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 189 of 211

190 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on September 24, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 190 of 211

191 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on October 2, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 191 of 211

192 Figure Five-day HYSPLIT back-trajectories initiated at 2200 UTC (1600 CST) on October 3, 2011 from Austin Northwest using starting heights Above Ground Level (AGL) of 10 meters (m), 500 m, and 1 km. Page 192 of 211

193 8.3 June 2012 TCEQ is currently in the process of developing a new photochemical modeling platform reflecting a June 2012 base case. Much of June 2012 was characterized by relatively low ozone levels in the Austin area, while other parts of eastern Texas experienced elevated ozone. Because the June 2012 episode is under consideration for photochemical modeling by TCEQ, it is useful to analyze synoptic weather patterns that may have contributed to widespread high ozone concentrations across the state, as well as investigate why the Austin area did not record as many high ozone events during this episode compared to other areas. Similar to the analysis of weather patterns during the August October 2006 and 2011 high ozone episodes summarized earlier in this section, AACOG investigated the large-scale atmospheric circulation features during Austin area high ozone episodes for this period using upper air and surface weather maps maintained by the Precipitation Diagnostics Group in the Mesoscale and Microscale Meteorology Division of NCAR. 34 Summaries were generated for each day of the episode with an emphasis on days that had Austin 8-hour ozone concentrations >= 70 ppb. The results are presented in Table 9-3 and include descriptive discussions of the relevant upper air and surface large-scale weather features in addition to local meteorological observational summaries at Austin Northwest (maximum daily temperature: T, daily average wind speed: WS, daytime wind direction: WD) and at Austin-Bergstrom International Airport (daily average relative humidity: RH). Also provided in Table 9-3 are the minimum and maximum 8-hour daily maximum ozone concentrations measured by the Austin monitoring network. The results presented in Table 9-3 are the same analyses that were performed for the representative high ozone episodes during (refer to Table 8-2). These results are provided so that a detailed and direct inter-comparison of conditions on specific June 2012 days can be made to those that occurred at other times during , if desired. Five-day HYSPLIT back trajectories for days with 8-hour ozone >= 70 ppb are not included in this section because they can be found in Section 8, Figures 8-5 through 8-8. Much of the month of June 2012 was characterized by persistent south to southeasterly surface flow that transported relatively clean maritime air into Central TX. These conditions prevented 8-hour ozone concentrations from exceeding 60 ppb between the 2nd and the 21st. There were two events where ozone exceeded 70 ppb: the first was at the beginning of the month and was short-lived, while the second was a more prolonged event between the 22nd and the 28th. Each of these events was preceded by the passage or proximity of a frontal boundary through the area. The first high ozone event was initiated by the passage of a front on May 31st. As seen with other high ozone episodes discussed earlier in this section, this resulted in the transport of drier continental air from the north. By June 2, SE flow off the Gulf resumed and rapidly moderated ozone levels in the area. There was another frontal boundary that passed through Central TX on June 8th. Although this weather pattern change did trigger an increase in 8-hour ozone concentrations, they remained between 50 and 60 ppb between the 5th and the 9th. The wind shift associated with the passage of the front was relatively short-lived, but stagnant surface conditions did exist both before and after the northerly wind shift. Other parts of Texas experienced moderate ozone levels (between 60 ppb and 75 ppb) during this time period, primarily in SE Texas and along the Interstate 35 corridor, with Houston and Dallas recording ozone above 75 ppb. On the 10th, steady SE flow resumed and persisted until the 22nd, when another front passed close to Central TX Page 193 of 211

194 During this period of prolonged SE surface winds, there was a brief period of elevated ozone over 50 ppb, which may have been associated with an approaching frontal boundary that dissipated before it passed through. Humidity levels and temperatures were slightly lower, but there was no pronounced wind shift. HYSPLIT surface and upper level back trajectories did indicate more easterly transport compared to southeasterly during the previous several days, although resultant wind direction observations from Austin Northwest did not indicate any change in wind direction to corroborate the HYSPLIT run. The final high ozone event in the June 2012 episode was initiated by the close approach of a frontal boundary. This began a week-long period of mostly stagnant air flow, with some days experiencing 180 degree wind shifts. A surface high pressure system over NE TX and Arkansas in conjunction with Tropical Storm Debby in the eastern Gulf resulted in weak steering flow over much of TX. Once Tropical Storm Debby moved offshore into the Atlantic, high pressure was allowed to build in the southeast U.S. and reestablish SE flow off the Gulf, which brought ozone levels back down to below 50 ppb. Page 194 of 211

195 Table 8-3. Daily local meteorological measurements at Austin Northwest and large-scale weather features for June 2012 Episode or Date 5/30/2012 Max Austin 8-Hr Ozone (ppb) and Site 39 (DS) 5/31/ (A) 6/1/ (A) 6/2/ (LG) Min Austin 8-Hr Ozone (ppb) and Site 32 (MR) 47 (SM) 61 (SM) 35 (MR) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD SSE 1.7 ESE E 1.7 NE NNE 1.3 NE S 1.6 SSE Discussion of Selected Large-Scale Weather Features Central TX is on the eastern edge of an upper level ridge, causing NW winds aloft. A frontal boundary is situated across Oklahoma and the TX Panhandle while a small area of surface high pressure is over the TX/LA border, providing SE flow into Central TX. The front clears Central TX by midday and erodes the eastern edge of the upper level ridge. Surface wind shifts occur in the mid morning hours. Showers and thunderstorms form just ahead of the boundary, primarily focused on NE TX. High pressure builds in behind the front, centered over the Great Plains and imparting N to NE winds at the surface, transporting continental air into Central TX. As the upper level trough associated with the frontal boundary moves east, NW winds continue aloft with Central TX located on the west side of the trough. Surface high pressure builds to the east as the front clears most of the country. This resumes SE flow into Central TX, moderating the elevated ozone from the previous day with clean maritime air from the Gulf of Mexico. Upper level trough associated with the cold front brings NW winds aloft. Page 195 of 211

196 Episode or Date 6/3/2012 6/4/2012 Max Austin 8-Hr Ozone (ppb) and Site 34 (DS) 46 (LG) 6/5/ (A) 6/6/2012 6/7/ (LG) 53 (NW) Min Austin 8-Hr Ozone (ppb) and Site 29 (NW) 35 (MR) 34 (MR) 45 (MR) 48 (MR) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD S 2.0 S S 0.8 SE 92.4* NNW 0.8* SSW* 93.4* * WSW* 1.7* S* NNW 1.5 NNE Discussion of Selected Large-Scale Weather Features Continued SE flow off of the Gulf further brings ozone levels down. High pressure aloft builds over the western Atlantic and Florida, stretching into the Gulf of Mexico. S to SE winds continue ahead of a stationary front draped across the TX/OK border. An upper level ridge builds over the Rockies, while a weak upper low develops over SW TX. Period of slightly elevated ozone begins as the frontal system slowly begins pushing south toward Central TX. A dry line passes through Central TX from the west, lowering humidity and enabling vertical mixing. A brief period of northerly winds in the morning hours that shifted to very weak southerly later on may have contributed to elevated ozone levels. A sharp ridge of high pressure aloft sits over the central U.S. Frontal boundary continues pushing slowly south as SW to SE flow persists ahead of it. Upper ridge still present over the central U.S., but another weak upper level low forms over SW TX. Frontal boundary clears Central TX and brings a wind shift out of the north. Upper ridge persists over the central U.S. Despite the frontal passage, humidity remains high, fueling clouds and showers across the state as the upper level low moves into central TX. A trace of rain is reported in Austin. Page 196 of 211

197 Episode or Date 6/8/2012 6/9/2012 6/10/2012 6/11/2012 Max Austin 8-Hr Ozone (ppb) and Site 59 (SM) 58 (LG) 44 (LG) 46 (LG) 6/12/ (A) 6/13/ (A/LG) Min Austin 8-Hr Ozone (ppb) and Site 49 (MR) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD NNE 2.7 NNE 52 (DS) SE 2.6 SSE 39 (NW) 35 (MR) 33 (MR) S 3.7 SSE S 3.9 S S 2.2 SSE 35 (H) NE 3.2 S Discussion of Selected Large-Scale Weather Features Rainfall continues across the state in the early part of the day as the frontal boundary sits over south TX and the upper low moves into southeast TX. Surface winds continue out of the north. SE surface winds return after a period of stagnant winds in the early morning. Upper level low begins moving off to the east. High pressure over the southeast becomes established as the front continues to move south. This resumes onshore flow off of the Gulf. A large E/W elongated upper level high pressure system develops across the SW part of the U.S. as the upper low continues to push off to the east. Southerly flow continues on the back side of a surface high pressure system over the SE U.S. A cool front moves into the TX Panhandle. High pressure aloft continues to strengthen across the southern U.S. Frontal boundary stalls out NW of the area. Surface S to SE winds continue. Weak upper level trough on the northern edge of the high pressure system is associated with the stalled frontal boundary. Variable winds in the morning become SE later in the day. Small high pressure system is just off the Gulf coast ahead of the stalled front. Weak surface low is over the TX Panhandle at the tail end of the front. Page 197 of 211

198 Episode or Date 6/14/2012 6/15/2012 6/16/2012 Max Austin 8-Hr Ozone (ppb) and Site 32 (A/DS) 36 (LG) 32 (A / DS / H) 6/17/ (H) 6/18/2012 6/19/2012 6/20/2012 6/21/ (LG) 28 (DS) 31 (DS) 51 (DS) Min Austin 8-Hr Ozone (ppb) and Site 21 (MR) 24 (MR) 21 (MR) 35 (MR) 39 (MR) 18 (MR) 26 (SM/H) 35 (MR) Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD S 3.6 SSE SSE 3.5 SSE SSE 2.5 SSE S 2.4 SE S 3.0 SSE SSE 3.6 SSE SE 4.2 SSE ESE 2.4 E Discussion of Selected Large-Scale Weather Features Continued SE surface flow with upper level high over southern U.S. Surface high pressure moves farther into the Gulf. Continued SE surface flow with upper level high over southern U.S. Continued SE surface flow with upper level high over southern U.S. Continued SE surface flow, but eastern half of the upper level ridge weakens. An upper level low develops over south TX and northern Mexico, providing NW flow aloft. Back trajectories indicate a more easterly origin of air flow at slower speeds than previous days. Background ozone levels appear to be higher, indicating transport may have occurred from industrial areas to the east. Surface flow continues to be out of the S to SE. Humidity and temperatures are both slightly lower than the previous few days. Continued SE surface flow with an area of high pressure aloft on either side of TX. Upper low moves south into northern Mexico and upper flow becomes more NE. Surface back trajectories continue to indicate more stagnant air flow into Central TX. Continued SE surface flow with an area of high pressure aloft on either side of TX. Frontal boundary enters the TX Panhandle. SE surface winds continue over Central TX. Page 198 of 211

199 Episode or Date 6/22/2012 Max Austin 8-Hr Ozone (ppb) and Site 64 (DS) 6/23/ (A) 6/24/ (A) 6/25/ (SM) Min Austin 8-Hr Ozone (ppb) and Site Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD 56 (LG) WNW 2.6 ESE 46 (MR) 38 (MR) 55 (MR) S 1.8 SE W 1.2 SE WNW 2.2 NE Discussion of Selected Large-Scale Weather Features Stationary front extends from a large low pressure system over Hudson Bay down into Northeast Texas. With a surface ridge at the tail end of the front, winds in central TX shift from generally westerly to easterly after noon. The area is on the eastern edge of an upper level high pressure system. Frontal presence over Texas moderates as the large low pressure over Hudson Bay weakens and lifts north. Surface high pressure becomes established over Arkansas, causing southerly flow across central TX. Upper level high moves east to become centered over the Texas Panhandle. Tropical Storm Debby forms in the eastern Gulf of Mexico. Upper level northeasterly flow with high pressure over the Panhandle and low pressure over the western Gulf. T.S. Debby strengthens slightly while moving NNE toward Florida. Surface high pressure remains centered over Arkansas but elongates farther west, creating a more southeasterly flow over central Texas, although winds are out of the west in the morning hours. Surface high pressure previously over Arkansas moves to the west to become centered over North Texas. T.S. Debby reaches peak intensity just offshore of the Florida Panhandle and imparts a weak northerly surface flow over central Texas before returning to southeasterly in the late afternoon. Page 199 of 211

200 Episode or Date 6/26/2012 6/27/2012 6/28/2012 6/29/2012 Max Austin 8-Hr Ozone (ppb) and Site 87 (NW) 81 (LG) 61 (A/LG) 46 (LG) Min Austin 8-Hr Ozone (ppb) and Site Daily Max T (F) Daily RH (%) 6a-10a WS (m/s) 6a-10a WD 12p-4p WS (m/s) 12p-4p WD 67 (H) W 0.8 ESE 72 (SM) SE 2.3 SSE 53 (H) S 2.4 S 36 (SM) S 3.6 SSE Discussion of Selected Large-Scale Weather Features Upper level high pressure over the Panhandle persists. Surface high pressure system over north Texas shifts west to the Rockies as a frontal system moves south, allowing development of seabreeze thunderstorms from the Gulf of Mexico to bring precipitation to south Texas. Surface winds during the morning are out of the SW and shift to E or SE later in the day. A surface high pressure system builds in over the southeastern U.S. as T.S. Debby moves away from Florida and into the Atlantic. This sets up a consistent southeasterly flow over central Texas after sunrise. Upper level high pressure reestablishes itself over the Great Plains, sending 100 temperatures as far north as Montana. High pressure over southeastern U.S. pushes southward to the Gulf of Mexico while an upper level low pressure develops over central Texas. Deep layer moisture flow from the Gulf of Mexico raises humidity levels across the state. High pressure over the SE U.S. and Gulf of Mexico continues SE flow from the Gulf and begins to moderate ozone levels. 6/30/ (A) 31 (H) SSE 3.3 SE Continued SE surface flow * Indicates that the data was collected from the Audubon monitor in the absence of data from the Austin Northwest monitor. Page 200 of 211

201 9 Relationship between Large-Scale Transport Conditions and the Austin Maximum Ozone As discussed in Section 7 of this report, high ozone episodes commonly occur following the passage of a cold front through Central Texas; typically, a ridge of high pressure moves south into Texas behind the front and the large-scale winds in the lower troposphere have a northeasterly or easterly component. Figure 9-1 shows the number of days >= 70 ppb when the maximum Austin-Round Rock MSA ozone concentration was measured at (1) Austin Northwest C3, (2) Audubon C38, (3) Dripping Springs C614, (3) Round Rock/Lake Georgetown/Hutto (these latter 3 monitors are all located to the north of Austin) C674, C690, and C6602, and (4) San Marcos/McKinney Roughs C675, C684, C1675 on high ozone days during In the event that two monitors had the maximum 8-hour ozone concentration in the region, the monitor with the highest 1-hour ozone concentration was chosen. Before 2010, Austin Northwest commonly reported the highest ozone concentration in the MSA monitoring network on high ozone days. Since 2010, there has been a relative increase in the number of maximum ozone days at monitors other than Austin Northwest, but a decline in the overall number of days with ozone concentrations >= 70 ppb since Figure 9-1. Number of days when a monitor measured the maximum region-wide 8-hour ozone maximum concentration when at least one measurement was >=75 ppb Although there is general similarity in the overall large-scale and local meteorological conditions during high ozone episodes, differences in the scale and evolution of the dominant weather features as well as the associated local atmospheric conditions such as temperature and wind speed make daily weather conditions unique. These differences are controlled by many factors, including the variability in overall global weather patterns that impact the dominant large-scale circulation features on both intra- and inter-annual time scales and drought conditions. Regardless of the specific weather conditions during a Page 201 of 211

202 given high ozone episode in the Austin area, it is reasonable to hypothesize that the maximum concentration in the local area would most often be located within the Austin urban plume (i.e., locallyformed ozone combined with background ozone entering Central Texas) that moves with the average winds in the lower troposphere. During periods with a varying wind direction or nearly calm conditions, a centrally-located monitor (e.g., Austin Northwest) might be expected to measure the maximum concentration, while conditions with moderate speeds and a relatively constant direction might be expected to transport the urban plume away from the urban core. To capture the large-scale wind conditions in the lower atmosphere during and prior to high ozone days in Central Texas, five-day HYSPLIT back-trajectories were initiated at the monitor that recorded the highest 8-hour ozone for that day using a starting height of 1 km AGL. The results were grouped by days that measured the maximum Austin concentration at Austin Northwest (Figure 9-2), Audubon (Figure 9-3), Dripping Springs (Figure 9-4), Round Rock/Lake Georgetown/Hutto (Figure 9-5), and San Marcos/McKinney Roughs (Figure 9-6). Austin Northwest has a wide variety of long-range inflow patterns to Central Texas (Figure 9-2). Of the thirteen cases where Austin Northwest recorded the highest 8-hour ozone, all but three occurred in the first half of the ozone season (April July). Additionally, the paths of the five-day back trajectories were generally shorter than for other monitors. Unlike the other monitors, no back trajectories originated in Canada and there were no cases of long-range transport from the northeast. Days when Austin Northwest recorded the highest 8-hour ozone generally had more stagnant and variable wind flow patterns. All but one of the four days when Audubon (northwest of the Austin urban core, Figure 9-3) measured the maximum concentration had long-range winds that had a northeasterly component, and three days occurred in the first half of ozone season (May June). On three of the days, this generally northeasterly wind flow pattern is consistent with the transport of a portion of the Austin urban plume towards the Audubon monitoring location, usually immediately following the passage of a cold front where prevailing winds are north or northeasterly. All ten days when Dripping Springs (west of the Austin urban core, Figure 9-4) measured the maximum concentration had long-range winds that had a northeasterly component, and all but one of the days occurred in the second half of ozone season (August October). The Dripping Springs back-trajectory paths within Texas are relatively smooth and straight-line suggesting a relatively steady-state (i.e., constant) wind direction in the lower troposphere. The exceptions to this pattern occurred only on the one day in the first half of the ozone season. This generally northeasterly wind flow pattern is consistent with the transport of a portion of the Austin urban plume towards the Dripping Spring monitoring location, usually immediately following the passage of a cold front where prevailing winds are north or northeasterly. In contrast, the back-trajectories for maximum ozone days that occurred at Round Rock/Lake Georgetown/Hutto monitors that are located close to each other on the north side of Austin (Figure 9-5) were consistent with northerly or northeasterly winds followed by an abrupt return flow from the south, suggesting possible recirculation of emissions from Texas (and local) sources followed by northward transport of the Austin urban plume to monitors located on the north side of the Austin area. Fifteen of the twenty-two back trajectories (68%) were consistent with this flow pattern. Of the seven back trajectories that did not follow this pattern, all but one indicated transport either from the north or east. The return flow from the south seen on the back trajectories is consistent with the passage of a front with high pressure building in behind it over the SE US. In Central TX, winds immediately following a Page 202 of 211

203 frontal passage are typically north or northeast, but gradually become southeasterly as high pressure builds in the SE US. Twelve of the sixteen high ozone days at San Marcos or McKinney Roughs (southeast and southwest of the Austin urban core, respectively) had back-trajectories consistent with northerly winds entering the Austin region throughout the lower atmosphere (Figure 9-6). Of the four remaining back trajectories, two showed transport from the southwest, indicating that the San Antonio urban plume may have impacted ozone at these monitors. All but three of the back trajectories for San Marcos/McKinney Roughs occurred in the second half of ozone season, and only one of those indicated long-range transport from the northeast. It is likely that these two monitors record the highest ozone immediately following the passage of a front when winds come from the north due to high pressure over the Great Plains. The results shown in Figures 9-2 through 9-6 suggest that specific large-scale transport patterns may be correlated with the location of the maximum monitored ozone concentration in the Austin area. Overall, these results suggest that the duration and consistency of lower tropospheric winds establish the overall movement of the Austin urban plume; the year-to-year (and episode-to-episode) variability in locations of the maximum Austin ozone is thus controlled by the spatial extent, evolution, and strength of the surface high pressure systems that establish the lower-atmospheric transport paths over Central Texas. Page 203 of 211

204 Figure 9-2. Inter-state back-trajectories (based on 5-day HYSPLIT back-trajectories initiated at 1 km AGL) on 13 days during that had the maximum Austin concentration at CAMS 3 (Austin Northwest). Page 204 of 211

205 Figure 9-3. Inter-state back-trajectories (based on 5-day HYSPLIT back-trajectories initiated at 1 km AGL) on 14 days during that had the maximum Austin concentration at CAMS 38 (Audubon). Page 205 of 211

206 Figure 9-4. Inter-state back-trajectories (based on 5-day HYSPLIT back-trajectories initiated at 1 km AGL) on 14 days during that had the maximum Austin concentration at Dripping Springs Page 206 of 211

207 Figure Inter-state back-trajectories (based on 5-day HYSPLIT back-trajectories initiated at 1 km AGL) on 22 days during that had the maximum Austin concentration at Round Rock, Lake Georgetown, or Hutto. Page 207 of 211

208 Figure 9-6. Inter-state back-trajectories (based on 5-day HYSPLIT back-trajectories initiated at 1 km AGL) on 16 days during that had the maximum Austin concentration at San Marcos or McKinney Roughs. Page 208 of 211