Air Dispersion Modeling Appendix

Transcription

1 Air Dispersion Modeling Appendix Prepared by: Lindsey Sears January 15, 2016

2 1. Introduction The Oxnard area of Ventura County is one of California s highest-producing strawberry regions. Many of the strawberry fields north of Oxnard lie close in proximity to neighborhoods such as Saticoy and schools in the Rio School District. I was asked to model impacts from the soil fumigants chloropicrin, 1-3 dichloropropene, and methyl bromide that were applied to various strawberry fields in The fields are located in the area surrounding Rio Mesa High School in north Oxnard. This report presents my modeling results and discusses the technical methodology I used for performing these analyses. Specifically, I prepared air concentrations and dose exposures of field fumigants, including chloropicrin, 1,3-dischloropropene (telone), and methyl bromide. These fumigants were applied to strawberry fields near Rio Mesa High Scholl during the period July 26, 2013 through August 3, Modeling Methodology This section describes the dispersion model, control options, and output options I used for modeling chloropicrin, 1,3-dichloropropene, and methyl bromide applied to strawberry fields. 2.1 Air Dispersion Model I performed this modeling analysis using USEPA s AERMOD air dispersion model, version 15181, obtained from the Support Center for Regulatory Atmospheric Modeling (SCRAM) website. Version is the latest version of the AERMOD model, which was publicly released on June 30, AERMOD Input Control Options I ran USEPA s AERMOD model with the following control options: Regulatory defaults Flagpole receptors Rural dispersion coefficients To correspond to a representative inhalation level, I used a flagpole height of 1.5 meters for all modeled receptors. I added this parameter to the receptor file when running AERMAP, as described in Section 3.4. I determined that this scenario should be modeled with the default AERMOD rural dispersion control option. I reached this finding using USEPA s methodology outlined in Section of

3 the Guideline on Air Quality Models Output Options My AERMOD modeling analyses of the Rio Mesa strawberry field fumigations incorporates 2013 site-specific meteorological data, corresponding with the application dates of the modeled chemicals. I calculated 168-hour (period) average air concentrations of the fumigants for determining cumulative exposure (in µg dose) and glutathione depletion potential (in mmoles of glutathione depletion). The output files generated by AERMOD provide the data necessary for preparing air concentration isopleths. 3. Model inputs The AERMOD air dispersion model requires a lengthy list of input values. Key inputs to this dispersion model include local geography, air emission rates of the released pollutant, source parameters (how and where the material is released to the air), receptors (locations where the offsite concentrations and deposition are calculated), and meteorological data (determines how and where the material is dispersed in the air). Each of these inputs is discussed below. 3.1 Geographical Inputs The ground floor of all air dispersion modeling analyses is establishing a coordinate system for identifying the geographical location of emission sources and receptors. These geographical locations are used to determine local characteristics (such as land use and elevation), and also to ascertain source to receptor distances and relationships. I used the Universal Transverse Mercator (UTM) NAD83 zone 11 coordinate system for identifying the easting (x) and northing (y) coordinates of the modeled fumigant sources and receptors. I verified the source coordinates using ESRI online high-resolution orthoimagery. As mentioned above, I determined that the emission sources should be modeled with rural dispersion coefficients. I studied the geographical setting in a three-kilometer radius circle surrounding the field locations, examining both land use and population density characteristics. If less than 50% of the surrounding area is urban and developed, then a rural classification is supported. Also, the default rural option may apply if the population density in the threekilometer radius surrounding each facility is less than 750 people per square kilometer. Since both of these conditions apply to the field locations, I modeled with AERMOD s default rural dispersion coefficients. 2 1 Id., Section USEPA, Revision to the Guideline on Air Quality Models: Adoption of a Preferred General Purpose (Flat and

4 3.2 Fumigant Source Parameters I modeled the fumigant air emissions using the AREAPOLY source type in AERMOD. The AREAPOLY source type is ideally suited to modeling irregularly shaped polygons, such as agricultural fields that release emissions from the ground surface. For the fumigant AREAPOLY sources, the following AERMOD inputs are required: A source identifier number or name; Source Location X (Easting) coordinate (UTM Zone 10, NAD83); Source Location Y (Northing) coordinate (UTM Zone 10, NAD83); Source base elevation (meters above sea level); Emission flux (g/(s-m 2 )); Release height of the area source (meters); Number of polygon vertices; X and Y coordinates for each polygon vertex (UTM Zone 11, NAD83); Initial vertical dispersion of the area source plume (meters). I applied a ground level release height (0 meters) and a vertical dimension of the plume equal to 10 feet (3.05 meters). From this vertical dimension I calculated an initial vertical plume dispersion input of 1.4 meters (SZINIT = 3.05 meters/2.15). The locations of the modeled field fumigations are shown in Annex Fumigant Flux Rates I reviewed pesticide use reports (PUR) and notices of intent (NOI) prepared by the fumigant applicants and submitted to the Ventura County Agricultural Commissioner. These documents show the fumigants applied to each field, as well as the quantity of pesticide, application method, date, and time, among other details. I used these reports to identify the fumigant emissions from the modeled strawberry fields. From these pesticide use reports I could calculate the amount of active ingredient (AI) of each pesticide applied, which is proportional to the quantity of emissions that volatilize off of each field. The California DPR has established emission ratings for field fumigants. An emission rating is described as: the emissions of fumigant to the air under field conditions, expressed as a proportion (percentage) of applied fumigant, and is fumigant- as well as application method- Complex Terrain) Dispersion Model and Other Revisions, Appendix W to 40 CFR Part 51, November 9, 2005, Section

5 specific. 3 An emission rating can be thought of as the total amount of fumigant expected to volatilize during and after the application. For methyl bromide, and the fumigation method used by the applicable strawberry growers, the emission rating is 48%; for chloropicrin, the emission rating is from 12 to 44% (depending on the field fumigation method); and for 1,3- dichloropropene, the emission rating is 19%. 4 Each field was fumigated on the dates listed in NOI. The PURs only listed the date and time that the last field (for each grower) was fumigated. For each field, I assumed that fumigations ended at the time listed in the PUR for the last field fumigated. I assumed that the fumigants volatilized in 168 hours after application. Specifically, I calculated AERMOD input flux rates (in µg/m 2 -hr) using the following assumptions: For methyl bromide, I used DPR's emission rating of 48% for TriCon 50/50 (48% of the applied MB volatilizes in 168 hrs). For chloropicrin, I used DPR's emission rating of 44% for Tri-Clor and TriCon 50/50 (44% of the applied CP volatilizes in 168 hrs). For dripped chloropicrin, I used DPR's emission rating of 12% for InLine (12% of the applied CP volatilizes in 168 hrs). For dripped 1,3-dichloropropene, I used DPR's emission rating of 19% for InLine (19% of the applied 1,3D volatilizes in 168 hrs). The pertinent fumigant application information from the notices of intent and pesticide use reports is detailed in Annex 2. The tables in this annex show the fumigant flux rate calculations over the anticipated emission release period of 168 hours following the fumigation in question. Each of the tables covers a specific NOI, which may incorporate multiple fields, multiple fumigants, and applications over a number of days. 3.4 Receptors I calculated air concentrations at a grid of receptors surrounding the fumigated fields and the community of Rio Mesa. The grid consists of receptors in an area large enough to develop the exposure contours (air concentrations and glutathione depletion) presented in the Annex 5 maps. 3 Terrell Barry, et al, Memo to John Sanders, Chief, Environmental Monitoring Branch, California DPR, Pesticide Volatile Organic Compound Emission Adjustments for Conditions and Estimated Volatile Organic Compound Reductions Revised Estimates, September 29, 2007, p California Department of Pesticide Regulation, Volatile Organic Compound Regulations, Fumigation Methods (FFM), FFM Codes for Pesticide Use Reporting, and Emission Ratings, December 2007, p. 1.

6 The receptor grid can be described as follows: 100 meter receptor spacing (UTM zone 11 NAD83): XUTM Range: to YUTM Range: to I created and modeled 7,384 gridded receptors for this analysis. As discussed earlier, I used a flagpole height of 1.5 meters (human inhalation level) for all modeled ground-level receptors. I also modeled a group of sensitive receptors in the Rio Mesa community area. The sensitive locations, with the associated UTM zone 11 coordinates are listed in the following table. The sensitive receptors were also modeled with a flagpole receptor height of 1.5 meters. Location UTMX UTMY Albert H. Soliz Library Del Mar Child Development Rio Vista Middle School Rio Del Valle Junior High School Rio Plaza Head Start Center Rio Real Elementary School Rio Mesa High School Rio Del Mar School Linda Vista Adventist Elementary School Linear Park Northbank Linear Park Riverview Linear Park Central Park Rodger Jones Community Park East Park Vineyard Park Kennebec Linear Park Modeled source and receptor locations require terrain elevation data, in meters above sea level. I obtained terrain elevation data for these locations using National Elevation Dataset (NED) GeoTiff data for the area encompassing the strawberry fields and the modeled receptors. GeoTiff is a binary file that includes data descriptors and geo-referencing information necessary for extracting terrain elevations. I extracted terrain elevations from the NED files using USEPA s AERMAP program, v , with 1/3 rd arc-second (10 meter horizontal) resolution.

7 3.5 Meteorological Data I prepared 2013 site-specific meteorological data to be used in my air dispersion modeling analysis. This year of data corresponds with the application dates of the modeled fumigants. I developed 2013 meteorological data that incorporates methods to reduce calm and missing hours (e.g. use one-minute data and USEPA s AERMINUTE program). 5 The meteorological data required by AERMOD is prepared by AERMET. Required data inputs to AERMET are: surface meteorological data, twice-daily soundings of upper air data, and the micrometeorological parameters surface roughness, albedo, and Bowen ratio. AERMET creates the model-ready surface and profile data files required by AERMOD. This section discusses how I prepared meteorological data to be used in my modeling analysis. Using AERMET v , I created an AERMOD-ready meteorological data set to model the fields. This data set covers 2013, and is summarized as follows: Meteorological data used for modeling the fields: Site-specific surface data: VCAPCD Rio Mesa High School. Supporting airport surface data: Camarillo Airport ASOS (KCMA); Regional upper air data: Vandenberg Air Force Base (KVBG); Site-specific surface data I obtained and processed hourly meteorological data collected at Rio Mesa High School. These data, collected by the Ventura County Air Pollution Control District (VCAPCD), are ideally suited for modeling the impacts of the strawberry field fumigations in the Rio Mesa community. These data were provided to me by the VCAPCD, and include hourly wind speed, wind direction, air temperature, relative humidity, and the standard deviation of the horizontal wind direction fluctuations (sigma-theta). The wind data are collected at an anemometer height of 10 meters. I processed these site-specific data using AERMET v , to provide the site-specific pathway data used for creating the AERMOD-ready surface and profile meteorological data sets Supporting airport surface data I used 2013 Integrated Surface Hourly (ISH) data obtained from the National Climatic Data Center (NCDC). From the ISH dataset, I extracted ASOS data from the Camarillo Airport. Based 5 USEPA, AERMINUTE User s Instructions, v , p. 1.

8 on distance and site characteristics, I determined that this airport is the most site-appropriate for modeling the strawberry fields. I also obtained 2013 one-minute ASOS wind data from the Camarillo Airport, which I processed with AERMINUTE v I downloaded these one-minute data from the NCDC. 6 I input the ice-free wind instrument start date (January 25, 2007) and used default settings with AERMINUTE. As a quality assurance measure, I compared values developed from the oneminute data with the corresponding ISH data file. The data from the KCMA ASOS site (ISHD) are used to supplement the site-specific data collected at Rio Mesa. The KCMA data are used to provide additional data necessary for AERMET to calculate certain AERMOD inputs, as well as for providing inputs in cases where wind measurements may be missing from the Rio Mesa data set. I processed the ISH data through AERMET Stage 1, which performs data extraction and quality control checks. I merged the AERMINUTE output files with the processed AERMET Stage 1 ISH and site-specific data, and the upper air pathway data in AERMET stage Upper Air Data I used 2013 upper air data from twice-daily radiosonde measurements obtained from Vandenberg Air Force Base. These data are in Forecast Systems Laboratory (FSL) format which I downloaded in ASCII text format from NOAA s FSL website. 7 I downloaded and processed all reporting levels with AERMET. Upper-air data are collected by a weather balloon that is released twice per day at selected locations. As the balloon is released, it rises through the atmosphere, and radios the data back to the surface. The measuring and transmitting device is known as either a radiosonde, or rawindsonde. Data collected and radioed back include: air pressure, height, temperature, dew point, wind speed, and wind direction. I processed the FSL upper air data through AERMET Stage 1, which performs data extraction and quality control checks AERSURFACE AERSURFACE is USEPA s program for extracting surface roughness, albedo, and daytime Bowen ratio for an area surrounding a given location. 8 AERSURFACE uses land use and land 6 See: ftp://ftp.ncdc.noaa.gov/pub/data/asos-onemin/ 7 Available at: 8 Albedo is the fraction of total incident solar radiation reflected by the surface back to space (whiter surfaces have higher albedo). The Bowen ratio is an indicator of surface moisture. It is the ratio of sensible heat flux to latent heat flux and drier areas have a higher Bowen ratio. Surface roughness, shown in shorthand as ( z 0 ), is an essential

9 cover (LULC) data in the U.S. Geological Survey s 1992 NLCD to extract the necessary micrometeorological data. I used these 1992 LULC data for processing meteorological data sets which then serve as input to AERMOD. I used AERSURFACE v to develop surface roughness, albedo, and daytime Bowen ratio values in a region surrounding the meteorological data collection site (Camarillo Airport). Using AERSURFACE, I extracted surface roughness in a one kilometer radius surrounding the data collection site. I also extracted Bowen ratio and albedo for a 10 kilometer by 10 kilometer area centered on the meteorological data collection site. I processed these micrometeorological data for seasonal periods using 30-degree sectors. I developed variable Bowen ratios, based on precipitation for each season of I determined the seasonal moisture conditions (wet, average, dry) using 1981 through 2010 climatic mean monthly rainfall data for the Camarillo Airport. 9 For each season, I compared the seasonal total rainfall to climatic means for that season. Seasonal rainfall less than 75% of climatic means was assessed as dry. I assessed seasonal rainfall greater than 125% of climatic means as wet. 10 Tables of the precipitation conditions for determining seasonal Bowen ratios from Camarillo Airport are included in Annex Data Review I did not fill missing hours in the meteorological data sets as the data files easily exceed USEPA s 90% data completeness requirement. 11 A wind rose of the AERMOD-ready meteorological data set I created for the 2013 Rio Mesa/Camarillo Airport/Vandenberg Air Force Base data set is included in Annex 3. parameter in estimating turbulence and diffusion. Technically, it s the height above the ground that the log wind law extrapolates to zero. For our purposes, z 0 can be thought of as a measure of how much the surface characteristics interfere with the wind flow. Very smooth surfaces, like short grass or calm ponds, have very low values of z 0 -- on the order of 0.01 meter or less. Tall and irregular surfaces, which are a greater obstacle to wind flow, have higher values of z 0 up to 1.0 meter or more for forests. 9 See 10 USEPA, Non-Hg Case Study Chronic Inhalation Risk Assessment for the Utility MACT Appropriate and Necessary Analysis, March 16, 2011, p USEPA, Meteorological Monitoring Guidance for Regulatory Modeling Applications, EPA-454/R-99-05, February 2000, Section 5.3.2, pp

10 4. Modeling Results As discussed above, I modeled air impacts from chloropicrin, 1,3-dichloropropene, and methyl bromide applied to various Oxnard strawberry fields in The modeled impacts are presented in Annex 4 and Annex 5, and discussed below. I summed the exposure from each fumigant at each receptor. The assumption is that the exposures from each field are cumulative, since the exposures are nearly simultaneous (there are some slight differences due to the dates that each field was fumigated). 4.1 Sensitive Receptors For a list of sensitive locations in the Rio Mesa community, I modeled 168-hour average air concentrations of chloropicrin, 1,3-dichloropropene, and methyl bromide following the field fumigations discussed above. I also calculated the intake (dose) of these fumigants using an adult inhalation rate of 20 cubic meters of air per day (140 cubic meters of air over 168 hours), and an intake to dose rate of 100%. Tables of these modeled air concentrations and doses at the sensitive receptors are included in Annex 4. At these sensitive receptors I also calculated glutathione (GSH) depletion using depletion factors provided to me. The GSH depletion factors I used are as follows: 2.43E-05 mmoles GSH consumed per microgram of chloropicrin 1.05E-05 mmoles GSH consumed per microgram of methyl bromide 9.01E-06 mmoles GSH consumed per microgram of 1,3-dichloropropene A table showing GSH depletion by fumigant, and for all fumigants combined, is also included in Annex Community Exposure Maps For the grid of receptors encompassing the Rio Mesa community, I modeled 168-hour average air concentrations of chloropicrin, 1,3-dichloropropene, and methyl bromide following the field fumigations discussed above. I also calculated the intake (dose) of these fumigants using an inhalation rate of 140 cubic meters of air over 168 hours, and an intake to dose rate of 100%. Maps of these modeled doses at the gridded receptors are included in Annex 5. A map showing GSH depletion for all fumigants combined is also included in Annex 5.

11 4.3 Glutathione Depletion Weighting Factors for the Three Fumigants This is the memo provided to Lindsey Sears for modeling, with weighting factors to evaluate potential overall glutathione depletion for exposures to the three fumigants. Exposures are stated in terms of micrograms and micrograms/cubic meter at each location and for each fumigant separately. Appropriate weighting factors to make these conversions depend on: The molecular weights (the number of moles of each fumigant per microgram is inversely proportional to the fumigant s molecular weight; each mole in turn represents a particular number of molecules about 6 X ). The number of molecules of glutathione potentially consumed per mole of exposure to each fumigant 4 in the case of chloropicrin, 2 in the case of dichloropropene, and 1 in the case of metam sodium according to the metabolism schema in Figure 2 on the next page. Therefore in numerical terms the weighting factors are: Chloropicrin 4 mmoles GSH/mmole fumigant. ( mg chloropicrin/mmole)* 1000 micrograms/mg = 2.43 X 10-5 mmoles GSH consumed per microgram of chloropicrin exposure Dichloropropene 1 mmoles GSH/mmole fumigant. ( mg dichloropropene/mmole)* 1000 micrograms/mg = 9.01 X 10-6 mmoles GSH consumed per microgram of dichloropropene Methyl isocyanate 1 mmole GSH/mmole fumigant. (73.12 mg methyl isocyanate/mmole)* 1000 micrograms/mg = 1.37 X 10-5 mmoles GSH consumed per microgram of methyl isocyanate

12 The latter calculation can also be done in terms of the weight of the parent metam sodium by adding back the weight of the sodium: Metam sodium 1 mmole GSH/mmole fumigant. ( mg metam sodium/mmole)* 1000 micrograms/mg = 7.74 X 10-6 mmoles GSH consumed per microgram of metam sodium Methyl bromide 1 mmole GSH/mmole fumigant. (94.95 mg methyl bromide/mmole)* 1000 micrograms/mg = X 10-5 mmoles GSH consumed per microgram of methyl bromide

13 Annex 1: Location Maps

14

15

16

17 Annex 2: Fumigation Flux Rate Calculations

18 NOI 1 Fumigant Flux Rates Application Period: July 26,2013 August 3, 2013 PUR Block- Modeled Modeled Vol Q Modeling Area Specific Area Area Period Rating Flux Rate File Date Time Fumigant (acre) AI (lb) Product (acres) (m 2 ) (hrs) (%) (g/(s-m 2 )) (.inp) July 26, :00 Methyl Bromide 1B TriCon 50/ E % 1.198E-05 M1B5 July 28, :00 Methyl Bromide 1B TriCon 50/ E % 1.198E-05 M1B6 July 30, :00 Methyl Bromide 1B TriCon 50/ E % 1.198E-05 M1B4 July 26, :00 Chloropicrin 1B TriCon 50/ E % 1.092E-05 C1B5 July 28, :00 Chloropicrin 1B TriCon 50/ E % 1.092E-05 C1B6 July 30, :00 Chloropicrin 1B TriCon 50/ E % 1.092E-05 C1B4 Total area (m 2 ): 3.47E+05 Methyl Bromide: total lbs AI applied total acres total acres Chloropicrin: total lbs AI applied 1,3-Dichloropropene: total lbs AI applied FFM: 1103 Q Rating MB: 0.48 Q Rating CP: 0.44 Q Rating 1,3-D:

19 NOI 2 Fumigant Flux Rates Application Period: July 26,2013 August 3, 2013 PUR Block- Modeled Modeled Vol Q Modeling Area Specific Area Area Period Rating Flux Rate File Date Time Fumigant (acre) AI (lb) Product (acres) (m 2 ) (hrs) (%) (g/(s-m 2 )) (.inp) July 29, :00 1,3-Dichloropropen 2A InLine E % 5.514E-06 D2A July 31, :00 1,3-Dichloropropen 2B InLine E % 5.514E-06 D2B August 1, :00 1,3-Dichloropropen 2C InLine E % 5.514E-06 D2C August 3, :00 1,3-Dichloropropen 2D InLine E % 5.514E-06 D2D July 29, :00 Chloropicrin 2A InLine E % 1.908E-06 C2A July 31, :00 Chloropicrin 2B InLine E % 1.908E-06 C2B August 1, :00 Chloropicrin 2C InLine E % 1.908E-06 C2C August 3, :00 Chloropicrin 2D InLine E % 1.908E-06 C2D Total area (m 2 ): 4.64E+05 Methyl Bromide: total lbs AI applied total acres total acres Chloropicrin: total lbs AI applied 1,3-Dichloropropene: total lbs AI applied FFM: 1209 Q Rating MB: Q Rating CP: 0.12 Q Rating 1,3-D: 0.19

20 NOI 3 Fumigant Flux Rates Application Period: July 26,2013 August 3, 2013 PUR Block- Modeled Modeled Vol Q Modeling Area Specific Area Area Period Rating Flux Rate File Date Time Fumigant (acre) AI (lb) Product (acres) (m 2 ) (hrs) (%) (g/(s-m 2 )) (.inp) July 26, :00 Chloropicrin 3NE Tri-Clor E % 1.654E-05 C3NE July 27, :00 Chloropicrin 3SE Tri-Clor E % 1.654E-05 C3SE July 28, :00 Chloropicrin 3SW Tri-Clor E % 1.654E-05 C3SW July 29, :00 Chloropicrin 3NW Tri-Clor E % 1.654E-05 C3NW Total area (m 2 ): 2.01E+05 Methyl Bromide: total lbs AI applied total acres total acres Chloropicrin: total lbs AI applied 1,3-Dichloropropene: total lbs AI applied FFM: 1103 Q Rating MB: Q Rating CP: 0.44 Q Rating 1,3-D:

21 NOI 4 Fumigant Flux Rates Application Period: July 26,2013 August 3, 2013 PUR Block- Modeled Modeled Vol Q Modeling Area Specific Area Area Period Rating Flux Rate File Date Time Fumigant (acre) AI (lb) Product (acres) (m 2 ) (hrs) (%) (g/(s-m 2 )) (.inp) July 28, :00 Chloropicrin 4N Tri-Clor E % 1.247E-05 C4N July 30, :00 Chloropicrin 4M Tri-Clor E % 1.247E-05 C4M August 1, :00 Chloropicrin 4S Tri-Clor E % 1.247E-05 C4S Total area (m 2 ): 3.85E+05 Methyl Bromide: total lbs AI applied total acres total acres Chloropicrin: total lbs AI applied 1,3-Dichloropropene: total lbs AI applied FFM: 1103 Q Rating MB: Q Rating CP: 0.44 Q Rating 1,3-D:

22 NOI 5 Fumigant Flux Rates Application Period: July 26,2013 August 3, 2013 PUR Block- Modeled Modeled Vol Q Modeling Area Specific Area Area Period Rating Flux Rate File Date Time Fumigant (acre) AI (lb) Product (acres) (m 2 ) (hrs) (%) (g/(s-m 2 )) (.inp) July 26, :00 1,3-Dichloropropen 5B InLine E % 5.454E-06 D5 July 26, :00 1,3-Dichloropropen 5B InLine E % 5.454E-06 D5 July 26, :00 1,3-Dichloropropen 5B InLine E % 5.454E-06 D5 July 26, :00 Chloropicrin 5B InLine E % 1.886E-06 C5 July 26, :00 Chloropicrin 5B InLine E % 1.886E-06 C5 July 26, :00 Chloropicrin 5B InLine E % 1.886E-06 C5 Total area (m 2 ): 1.60E+05 Methyl Bromide: total lbs AI applied total acres total acres Chloropicrin: total lbs AI applied 1,3-Dichloropropene: total lbs AI applied FFM: 1209 Q Rating MB: Q Rating CP: 0.12 Q Rating 1,3-D: 0.19

23 Annex 3: Meteorological Data Supporting Information

24 Attachment 2: Precipitation Conditions for Determining Seasonal Bowen Ratios (Based on 1981 through 2010 Precipitation Data from Oxnard Weather Forecast Office (USC )) Year Winter Dec-Feb Spring Mar-May Summer Jun-Aug Fall Sep-Nov Year Jan-Dec Winter Dec-Feb Spring Mar-May Summer Jun-Aug Fall Sep-Nov Year Jan-Dec Dry Average Dry Average Average Dry Wet Dry Wet Wet Wet Wet Wet Wet Wet Dry Dry Wet Wet Dry Dry Dry Dry Wet Dry Average Average Dry Wet Average Dry Dry Wet Average Dry Dry Average Dry Dry Dry Dry Dry Dry Average Dry Dry Dry Dry Dry Dry Average Wet Wet Dry Average Wet Wet Wet Dry Wet Wet Dry Wet Dry Wet Dry Dry Dry Dry Dry Wet Wet Dry Dry Wet Wet Average Dry Wet Wet Average Dry Average Average Average Wet Wet Dry Dry Wet Dry Wet Wet Dry Dry Average Average Dry Dry Average Wet Wet Dry Wet Wet Dry Dry Dry Wet Dry Average Wet Dry Dry Average Average Dry Dry Wet Average Wet Average Dry Dry Wet Dry Wet Dry Dry Average Dry Dry Dry Dry Dry Wet Dry Dry Average Average Average Dry Wet Dry Dry Wet Dry Dry Wet Wet Dry Wet Wet Wet Average Dry Average Dry Dry Dry Dry Dry Dry Dry Dry Average Dry Dry Dry Dry Dry Dry Dry Dry Dry Averages: Precipitation units: inches Average precipitation based on data

25 WIND ROSE PLOT: VCAPCD Rio Mesa met data; KCMA supplemental Upper air: KVBG (Vandenberg AFB) NORTH 30% 24% 18% 12% 6% WEST EAST WIND SPEED (Knots) >= SOUTH Calms: 0.00% COMMENTS: AERMET v ; AERMINUTE v DATA PERIOD: 2013 Jan 1 - Dec 31 00:00-23:00 CALM WINDS: 0.00% TOTAL COUNT: 8760 hrs. AVG. WIND SPEED: PROJECT NO.: 5.24 Knots WRPLOT View - Lakes Environmental Software

26 Annex 4: Exposure Tables for Sensitive Receptor Locations

27 Rio Mesa Sensitive Receptors Modeled Chloropicrin Air Concentrations (µg/m 3 ) and 168 hour total dose (µg) Location UTMX UTMY C1B4 C1B5 C1B6 C2A Albert H. Soliz Library Del Mar Child Development Rio VIsta Middle School Rio Del Valle Junior High School Rio Plaza Head Start Center Rio Real Elementary School Rio Mesa High School Rio Del Mar School Linda Vista Adventist Elementary School Linear Park Northbank Linear Park Riverview Linear Park Central Park Rodger Jones Community Park East Park Vineyard Park Kennebec Linear Park C2B C2C C2D C3NE C3NW C3SE C3SW C4M C4N C4S C5 Total Dose (µg)

28 Rio Mesa Sensitive Receptors Modeled 1,3 Dichloropropene Air Concentrations (µg/m 3 ) and 168 hour total dose (µg) Location UTMX UTMY D2A D2B D2C D2D D5 Total Dose (µg) Albert H. Soliz Library Del Mar Child Development Rio VIsta Middle School Rio Del Valle Junior High School Rio Plaza Head Start Center Rio Real Elementary School Rio Mesa High School Rio Del Mar School Linda Vista Adventist Elementary School Linear Park Northbank Linear Park Riverview Linear Park Central Park Rodger Jones Community Park East Park Vineyard Park Kennebec Linear Park

29 Rio Mesa Sensitive Receptors Modeled Methyl Bromide Air Concentrations (µg/m 3 ) and 168 hour total dose (µg) Location UTMX UTMY M1B4 M1B5 M1B6 Total Dose (µg) Albert H. Soliz Library Del Mar Child Development Rio VIsta Middle School Rio Del Valle Junior High School Rio Plaza Head Start Center Rio Real Elementary School Rio Mesa High School Rio Del Mar School Linda Vista Adventist Elementary School Linear Park Northbank Linear Park Riverview Linear Park Central Park Rodger Jones Community Park East Park Vineyard Park Kennebec Linear Park

30 Rio Mesa Sensitive Receptors Glutathione (GSH) Depletion (mmoles) Location UTMX UTMY Chloropicrin Dichloropropene Methyl Bromide Total Albert H. Soliz Library Del Mar Child Development Rio VIsta Middle School Rio Del Valle Junior High School Rio Plaza Head Start Center Rio Real Elementary School Rio Mesa High School Rio Del Mar School Linda Vista Adventist Elementary School Linear Park Northbank Linear Park Riverview Linear Park Central Park Rodger Jones Community Park East Park Vineyard Park Kennebec Linear Park

31 Annex 5: Community Exposure Maps

32

33

34

35