Elizabeth C. Weatherhead, PhD University of Colorado at Boulder. April, 2005 Revised October, Report to the U.S. Environmental Protection Agency

Transcription

1 Task (c) Report, Contract 4D-5888-WTSA: Submitted for APM 227 Report on Geographic and Seasonal Variability of Solar UV Radiation Affecting Human and Ecological Health Elizabeth C. Weatherhead, PhD University of Colorado at Boulder April, 25 Revised October, 26 Report to the U.S. Environmental Protection Agency Task (c) Analyze the major factors affecting the UV-B flux, including solar angle, latitude, elevation, cloud cover, pollution levels and composition, etc. These results have been peer-reviewed through a process supervised by EPA. Cooperative Institute for Research in the Environmental Sciences University of Colorado at Boulder 325 Broadway Boulder, CO 835 betsy.weatherhead@colorado.edu +1 (33)

2 EXECUTIVE SUMMARY The U.S. Environmental Protection Agency s Ultraviolet Monitoring Program began collecting data in 1995 using Brewer instruments at various national park and urban sites. The dates of the Brewer network operation correspond with an interesting time in terms of changes in total column ozone and other parameters, including surface pollutants. In this report, the relative influences of solar zenith angle, clouds/aerosol, total column ozone, and atmospheric pollution on surface UV variability are examined using data from the Brewer network. The effects of pollutants on UV reaching Earth s surface were examined using radiative transfer models to simulate the reduction in UV for a large range of surface ozone and sulfur dioxide levels. The results indicate that for the levels of surface ozone and sulfur dioxide pollution typically observed in the U.S., the UV radiation reaching the surface tended to be reduced by less than a few percent. Full spectral information from the Brewer network provided the ability to assess the effects of solar zenith angle, clouds, and total column ozone on surface UV. Daily ozone values for the analysis were determined by the A-pair ratio of wavelengths 313 nm to 331 nm, and UVA radiation measured at 34 nm at 6 degrees solar zenith angle was used to estimate the cloudiness for each day. These ratios were compared to total column ozone measurements from satellite (SBUV/2) and showed extremely high correlations, often greater than.9. The importance of the solar zenith angle, clouds and total column ozone was assessed using a multi-linear regression approach to incorporate the effects of solar zenith angle, clouds, and total column ozone. The results indicate that the largest single factor affecting biologically weighted daily UV doses is solar zenith angle, followed by clouds/aerosol, and then total column ozone. The percent of UV variability explained by solar zenith angle ranged from 37 percent at Hawaii Volcanoes National Park to 84 percent at Albuquerque. The total amount of variability explained by Sun angle, clouds/aerosol, and total column ozone in combination ranged from 77 percent to 91 percent for all stations with exception of stations at Virgin Islands, Hawaii, and Everglades. For these low latitude stations, the total explained variability was between 68 and 69 percent. The dependence of UV on latitude and elevation were quantified using both clear sky and all sky UV data. The results suggest a near linear decrease in both the clear sky and all sky UV irradiance with an increase in latitude and a less clear increase in UV with elevation, which does not appear to be linear. The high level of explained variability in the Brewer UV values for the 18 mid- and high latitude sites provides strong testimony of the quality of the screened measurements and the overall usefulness of the network. 2

3 TABLE OF CONTENTS 1. Introduction and Background Methods Quality Assurance Analysis of Factors Affecting UV Partitioning the Effects Results for Each of the 21 Brewer Locations Results for the Entire Brewer Network UV Variability as a Function of Latitude UV Variability as a Function of Elevation Conclusion References

4 1. Introduction and Background Task (c) analyze the major factors affecting the UV-B flux, including solar angle, latitude, elevation, cloud cover, pollution levels and composition, etc. is one of several tasks defined by the U.S. Environmental Protection Agency (EPA) to analyze measurements from the network of ultraviolet radiation instruments operated by EPA and the University of Georgia (UGA) in collaboration with the National Park Service (NPS). Other tasks address different elements of UV-B exposure and effects, and the results of those tasks are summarized in the individual task reports. The tasks defined by the U.S. Environmental Protection Agency at the start of the project are as follows: (a) Determine the trends in UV-B flux at the individual Brewer sites, at groups of similar sites and/or across the network. (b) Analyze the factors affecting the observations and trends at each site (and across the network, as appropriate) including correlations with changes in the ozone column, changes in stratospheric ozone, changes in ground level ozone, changes attributable to other pollutants or atmospheric constituents, etc. (c) Analyze the major factors affecting the UV-B flux, including solar angle, latitude, elevation, cloud cover, pollution levels and composition, etc. (d) Analyze the direct versus indirect exposures to UV-B radiation and the factors affecting the ratio. (e) Compare the UV-B flux measurements with the predictions of the National Weather Service Ozone Watch Program, together with an analysis of the differences and potential causes for the discrepancies. (f) Compare the Brewer UV-B measurements with those of the Tropospheric Ultraviolet (TUV) Model, Total Ozone Mapping Spectrometer (TOMS) measurements, and provide an analysis of the differences of those measurements and the causes of those differences. (g) Compare the UV-B measurements with air pollution measurements to determine the effects of tropospheric air pollution on UV-B exposures. (h) Compare trends in UV-B flux measured by the network at mid-latitudes in the United States to United National Environmental Program (UNEP) data. (i) Determine the effect of clouds/haze/aerosols on UV-B exposure. 4

5 (j) Analyze the directional diffuse (cloudless) sky irradiance in the nm (UV- B and 325-4nm (UV-A) wavelength bands as a function of aerosol optical depth. (k) Analyze the directional diffuse sky irradiance in the 29-4nm (UV-B and UV- A) wavelength band as a function of cloud cover, cloud type, cloud depth. (l) Analyze the reflectance (spectral albedo) for key materials (snow, beach sand, concrete, asphalt and water) in the 29nm-4nm (UV-B and UV-A) wavelength band, as appropriate for the network measurement sites. (m) Analyze the bi-directional reflectance (forward-scattering and back-scattering) for key materials (snow, beach sand, concrete, asphalt and water) in the 29nm 4nm (UV-B and UV-A) wavelength band, as appropriate for the network. Task (c) examines the major factors influencing the UV-B flux. These factors include clouds, pollution, and other atmospheric parameters, as well as latitude, elevation, and solar zenith angle. Of these factors, solar zenith angle is recognized to have the strongest effect on surface UV-B flux. Clouds are also a strong factor; compared to cloud cover effects, other atmospheric parameters play only a secondary role. Each of these factors is discussed in greater detail below. The EPA Brewer network is ideally designed to look for the effects of these various factors because of both the instrument and network design. Recording UV radiation at specific sun angles, as well as at solar noon, allows for comparison of measurements both at a particular site and at across the network while minimizing the impact of sun angle. The network design included measurements at a variety of sites and altitudes allowing for direct assessment of factors, such as altitude, in a systematic manner. The recording of measurements at every half nanometer will greatly assist in distinguishing the effects of ozone from the effects of clouds. Solar zenith angle. Solar zenith angle (SZA) is the primary factor affecting UV-B flux. The solar zenith angle is the angle between the elevation of the Sun above the horizon and the local zenith, which is directly overhead. If the Sun is directly on the horizon, for instance, the solar zenith angle would be approximately 9 degrees. If the Sun were directly overhead, as for noontime in the tropics, the solar zenith angle would be near zero. Solar zenith angle changes with time of day as the Sun rises and sets in the sky. It also changes with time of year: in the northern mid-latitudes, noontime solar zenith angles are larger in winter than they are in summer because the Sun is farther south and therefore lower in the sky. The changes in SZA with time of day and time of year are responsible for the observed, often large, diurnal and seasonal changes in the UV-B flux at each site. SZA also depends on the site latitude, with more northern sites experiencing larger SZAs and thus lower levels of direct UV-B. 5

6 Clouds. Of the various atmospheric constituents affecting UV, clouds have the greatest effect on the UV-B flux reaching the surface. Under overcast conditions, clouds cause an overall decrease in surface UV amounts (Josefsson and Landelius, 2; Renaud et al., 2). However, broken cloud conditions can actually enhance UV amounts at the surface (Sabburg and Wong, 2; Weihs et al., 2; Estupiñán et al., 1996), with the most important being whether or not the Sun is obscured (Grant and Heisler, 2; Schwander et al., 22). Because cloud conditions can change rapidly over the course of a day, or even within the time period of a Brewer scan, quantifying the effects of clouds on the UV-B flux can be a difficult process. However, the fact that the Brewer measures as far into the UVA part of the spectrum as 363 nm allows us to identify cloud/aerosol effects with minimal influences from either ozone or sulfur dioxide. Pollutants. Ozone, sulfur dioxide, and other pollutants in the lower atmosphere can absorb UV-B radiation and prevent a portion of it from reaching the surface. Absorbing aerosols and gaseous species also decrease the UV-B flux. Various investigators (e.g., McArther et al, 1999; Ilyas et al., 21; Kirchoff et al., 21; Jacobson, 21; di Sarra et al., 22) have assessed the UV effects of aerosol from biomass burning, forest fires, desert dust, and anthropogenic sources. Increased air pollution levels have been observed to decrease surface erythemal UV by as much as 3 to 4 percent (Papayannis et al., 1998; Repapis et al., 1998). However, in general the effects of pollutants tend to be much smaller in magnitude than the UV impacts from changing cloud or ozone amounts. The effects of pollutants are assessed using the largest values ever likely to be seen in the network, as well as the typical levels seen. Because of the Clean Air Act and its amendments, the levels observed in the U.S. are much less than those observed in developing countries. The near-surface levels of sulfur dioxide and ozone are particularly low compared to those in developing countries. 3. Methods Quality Assurance Before the factors affecting the UV-B flux could be assessed, the UV data required evaluation for quality. Dr. Weatherhead conducted a thorough review of the available data with the scientists at the University of Georgia (UGA) to determine any potential biases at particular sites. The UGA personnel have put considerable effort into producing a Level 1 data product suitable for scientific use. The Level 1 data from all 21 sites were used to address several questions about the various factors influencing UV. The spectral nature of the Brewer data and large geographic coverage of the Brewer network (Figure 1) provide important information for assessing the factors affecting UV throughout the continental United States and in Alaska and Hawaii. 6

7 Denali Denali Olympic Glacier Theodore Acadia Rockymtn Boulder Sequoia Canyonlands Albuquerque Riverside Chicago Gaithersburg Shenandoah Greatsmoky Rtp Atlanta Bigbend Everglades ii Hawaii Virginislands Figure 1. The locations of the instrument sites for the EPA Brewer UV monitoring network operated by the University of Georgia are shown. The sites cover a large geographical area and represent a range of ecosystems. Parameters relating to UV observations and temperature characterizations at each of the Brewer sites are summarized in Table 1. Columns a through c give the latitude, longitude, elevation, and year observations were begun at each site. 7

8 Table 1. Location and other information for the 21 EPA Brewer sites. Site (Brewer #) (a) Latitude/ Longitude (b) Elevation (c) Start Date Glacier National Park, 48.7ºN, 113.4ºW 424 m 1997 Montana (96-134) Denali National Park, 63.7ºN, 149.ºW 661 m 1997 Alaska (96-141) Olympic National Park, 48.1ºN, 123.4ºW 32 m 1997 Washington (96-147) Rocky Mountain National 4.ºN, 15.5ºW 2896 m 1998 Park, Colorado (96-146) Hawaii Volcanoes National 19.4ºN, 155.3ºW 1243 m 1999 Park, Hawaii Boulder, CO (93-11) 4.1ºN, 15.2ºW 1689 m 1996 Gaithersburg, MD (15) 39.1ºN, 77.2ºW 43 m 1994 Acadia National Park, 44.4ºN, 68.3ºW 137 m 1998 Maine (96-138)* Everglades National Park, 25.4ºN, 8.7ºW 18 m 1997 Florida (96-135) * Chicago, IL (94-13) 41.8ºN, 87.6ºW 165 m 1999 Atlanta, GA (94-18) * 33.8ºN, 84.4ºW 91 m 1994 Research Triangle Park, 35.9ºN, 78.9ºW 14 m 1995 NC (92-87) Great Smoky National 35.6ºN, 83.8ºW 564 m 1996 Park, TN (96-132) * Big Bend National Park, 29.3ºN, 13.2ºW 329 m 1997 Texas (96-13) Albuquerque, NM (94-19) 35.1ºN, 16.6ºW 1615 m 1998 * Sequoia National Park, 36.5ºN, 118.8ºW 549 m 1998 California (96-139) * Virgin Islands National 18.3ºN, 64.8ºW 3 m 1998 Park, U.S. Virgin Islands (96-144) Shenandoah National Park, 38.5ºN, 78.4ºW 325 m 1997 VA (96-137) Canyonlands National 38.5ºN, 19.8ºW 814 m 1997 Park, Utah (96-133) Riverside, CA (94-112) 34.ºN, 117.3ºW 84 m 1995 Theodore Roosevelt National Park, North Dakota (96-131) * 46.9ºN, 13.4ºW 238 m 1998 *A site that has a shift or other discontinuity Problem values in the data, including missing scans or extremely high or low points, were originally flagged in the Level 1 data files released by UGA. However, examination of the data has identified suspicious values that had not been flagged. The occurrence of such values varies from site to site, with some sites achieving extremely good collection rates as well as high data quality. The results of extensive quality screening of the data 8

9 are summarized in Figure 2. For each day, the data shown in red have been evaluated as being problem-free, based on set criteria. Over 35, days of data have been identified as good: each of these good days contains between 1 and 5 UV scans. Acadia : 1693 good days FLAGS - IDENTIFIED AS GOOD DAYS Albuquerque : 1685 good days Atlanta : 184 good days Bigbend : 1425 good days Boulder : 185 good days Canyonlands : 1889 good days Chicago : 1315 good days Denali : 112 good days Everglades : 1664 good days Gaithersburg : 239 good days Glacier : 1981 good days Great Smoky : 1797 good days Haw aii : 1154 good days Olympic : 1899 good days R.T.P : 171 good days Riverside : 2328 good days Rocky Mountain : 1674 good days Sequoia : 1626 good days Shenandoah : 227 good days T. Roosevelt : 1532 good days Virgin Islands : 1356 good days Weatherhead Sun Sep 26 14:43:3 UMS 24 task1.ssc 9

10 Figure 2. The data presented above are as released on the University of Georgia web site and screened for quality. These data are used in this report to examine spectral UV changes over the time periods of instrument operation. The UGA staff screened the data based on the two criteria listed in the readme file associated with the site. The two criteria suggested for screening were that the data compared reasonably well to a clear sky model and that approximately the right number of scans were taken on a given day. In addition to the screening performed by UGA as discussed above, we employed at least seven additional criteria. These additional criteria eliminated roughly 12% of the daily values. The checks were necessary to prevent systematic problems which could have had large impacts on the results presented. However, detailed screening of the spectral data was not possible and it is likely that some critical problems still exist in the individual spectral values. The additional quality assurance tests are summarized as follows: Test 1: The data were checked to define the time of the first scan. For the daily data to pass this test, the first scan had to have occurred within one hour of when the first scan was scheduled to occur. Test 2: The data were checked to define the time of the last scan. For the daily data to pass this test, the last scan had to have occurred within one hour of when the last scan was scheduled to occur. Test 3: The minimum solar zenith angles were checked. This test verifies whether scans were taken near solar noon. For the daily data to pass this test, the minimum solar zenith angle recorded was required be within three degrees of the expected solar zenith angle for that time of year. Test 4: The number of scans in a single day was checked. This test evaluates whether enough data were taken to provide a reasonable estimate of the daytime-integrated value. The criterion for this test is that the number of scans must not differ from the scheduled number of scans for the time of year by more than 1. Test 5: The data were visually evaluated for spikes. Spikes are a well-known phenomenon with the Brewer instruments, though their causes are not completely understood. Spikes result in extremely large values for single wavelength measurement in a single scan and do not represent a robust measurement of atmospheric composition. Many spikes had already been removed through UGA s screening of the data. 1

11 Test 6: The data were assessed for problems with inappropriate time sequences. Occasionally, the data were written to files incorrectly, with two days of data recorded to one daily file. In these cases, the daytime-integrated UV was observed to be much higher than usual. We identified and screened these instances by looking for cases in which the order of the solar zenith angles was out of line with what would happen during a normal day. The test also looked for uneven or highly irregular sampling during a day. Test 7: The data were checked to assure that DUV values were within range. Observed daytime-integrated UV values are rarely above 7 Joules/meter 2 /day and should never be equal to zero, given the sensitivity and locations of the Brewer instruments. All reported values greater than 9 and less than or equal to zero were omitted from further evaluation. No efforts were made to determine the cause of these unusual values, however an inspection of when they occurred at each site showed no patterns consistent with an atmospheric cause of the unusual values. Data were required to pass all seven tests to be used in the analysis of the Diffeyweighted UV. The number of data points removed in the screening process represented only a small portion of the data collected (less than 15% network wide), but the screening was necessary because only a few unusual points have the ability to change the results of the analyses. Analysis of Factors Affecting UV In this analysis, the UV-B flux values were analyzed in combination with radiative transfer models to assess the effects of solar zenith angle, site latitude and elevation, and clouds and atmospheric pollutants on the UV-B flux. The year-round monitoring and spectral coverage of the Brewer instruments, as well as the network s geographic coverage, provide the majority of the information required to complete these tasks. The focus of this report will be on the CIE daytime-integrated UV levels because of their importance to biological and material impacts. Analysis of these factors with respect to spectral trends is addressed in the Task 2 report. The minimum solar zenith angle as a function of time of year is expressed in Figure 3. This plot shows graphically how both the maximum solar zenith angle and the change of zenith angle with time of year vary from station to station. Stations near or within the tropics, including Hawaii and Virgin Islands, have peak sun angles twice each year, while mid-latitude locations peak only at the summer solstice. 11

12 Acadia Albuquerque Atlanta Bigbend Boulder Canyonlands Chicago Denali Everglades Gaithersburg Glacier Great Smoky Hawaii Olympic R.T.P Riverside Rocky Mountain Sequoia Shenandoah T. Roosevelt Virgin Islands Minimum Solar Zenith Angle D N O S A J J M A M F J Figure 3. Variations in the minimum solar zenith angle with time of year for each Brewer site. The effect of solar zenith angle on UV radiation is notable in the next seven pages of plots. These plots show the daily CIE-weighted UV for each location as a function of solar zenith angle. 12

13 Acadia 9th Percentile MEDIAN MEAN Cosine Minimum Solar Zenith Angle th Percentile MEDIAN MEAN Albuquerque Cosine Minimum Solar Zenith Angle Atlanta 9th Percentile MEDIAN MEAN Cosine Minimum Solar Zenith Angle 13

14 Bigbend 9th Percentile MEDIAN MEAN Cosine Minimum Solar Zenith Angle Boulder 9th Percentile MEDIAN MEAN Cosine Minimum Solar Zenith Angle th Percentile MEDIAN MEAN Canyonlands Cosine Minimum Solar Zenith Angle 14

15 Chicago 9th Percentile MEDIAN MEAN Cosine Minimum Solar Zenith Angle Denali 9th Percentile MEDIAN MEAN Cosine Minimum Solar Zenith Angle th Percentile MEDIAN MEAN Everglades Cosine Minimum Solar Zenith Angle 15

16 th Percentile MEDIAN MEAN Gaithersburg Cosine Minimum Solar Zenith Angle Glacier 9th Percentile MEDIAN MEAN Cosine Minimum Solar Zenith Angle th Percentile MEDIAN MEAN Great Smoky Cosine Minimum Solar Zenith Angle 16

17 Hawaii 9th Percentile MEDIAN MEAN Cosine Minimum Solar Zenith Angle Olympic 9th Percentile MEDIAN MEAN Cosine Minimum Solar Zenith Angle R.T.P 9th Percentile MEDIAN MEAN Cosine Minimum Solar Zenith Angle 17

18 Riverside 9th Percentile MEDIAN MEAN Cosine Minimum Solar Zenith Angle th Percentile MEDIAN MEAN Rocky Mountain Cosine Minimum Solar Zenith Angle Sequoia 9th Percentile MEDIAN MEAN Cosine Minimum Solar Zenith Angle 18

19 th Percentile MEDIAN MEAN Shenandoah Cosine Minimum Solar Zenith Angle th Percentile MEDIAN MEAN T. Roosevelt Cosine Minimum Solar Zenith Angle th Percentile MEDIAN MEAN Virgin Islands Cosine Minimum Solar Zenith Angle 19

20 The flux of UV radiation at the top of the atmosphere is proportional to the cosine of the instantaneous solar zenith angle. UV radiation at the bottom of the atmosphere, as presented in the previous plots, does not follow this simple rule for two reasons. First, the attenuation of UV as it passes through the atmosphere is stronger for low sun angles (high solar zenith angles). Second, the daytime-integrated UV values involve integration of UV over the course of a day, thereby including a full range of solar zenith angles from ninety degrees to the minimum reached at solar noon. Nonetheless, it is clear from these plots that sun angle has a large influence on the UV radiation measured at the EPA Brewer sites. These plots can be used to define clear sky UV radiation. For the purposes of these analyses, clear sky UV for each site is defined as the value of daytime-integrated, CIEweighted UV that is the cut-off for the top 5% of the values at a given minimum solar zenith angle. This distinction results in clear sky UV being defined independent of season. Because total column ozone in the northern hemisphere peaks in the springtime and has a low in the autumn, this definition of clear sky UV is not highly dependent on total column ozone levels. As an additional note, it is clear from these plots that the quality assurance and screening applied both by the University of Georgia and by Global Environmental Engineering Consultants, Inc. has resulted in a dataset that appears robust and free of gross errors. This robustness is impressive given the complexity of UV radiation measurements and the difficulty in properly interpreting them. Pollutant Effects The effects of pollutants, including surface ozone and SO 2, are assessed using a radiative model. The model used in this task implements the Discrete-Ordinate-Method Radiative Transfer model (DISORT) developed by Knut Stamnes (Stamnes et al., 1988). DISORT is a multi-stream radiative transfer model able to quantify the transfer of radiation in a scattering and emitting plane-parallel atmosphere. An interface improving the model s suitable for UV applications was added by Sasha Madronich at the National Center for Atmospheric Research, and improved further by Nataly Chubarova at Moscow State University in Russia. The model uses a reference Rayleigh profile to simulate the molecular atmosphere and provides output in 8 vertical layers at a resolution of 1 km. The standard atmosphere profile is the 1976 reference atmosphere, and ozone absorption coefficients are those developed by Molina and Molina (1986) for temperatures of 263 K and 298 K. Sulfur dioxide (SO 2 ) values are from McGee and Burris and nitrogen dioxide (NO 2 ) is based on Davidson. For our purposes, pollution is concentrated in the lowest 1 km at levels 5% of ground concentration. 2

21 Q_CIE in W/m2(eff) In coordination with Nataly Chubarova, the model was run to quantify the sensitivity of the UV-B flux to changes in atmospheric pollutants. The default values used in the study were as follows: Latitude 4º N Solar zenith angle 6º Elevation sea level Cloudiness % Pollution varied* Ozone 34 DU The latitude, solar zenith angle, elevation, and total column ozone amounts represent typical levels found in the EPA Brewer network. The pollution levels have been varied to represent a wide range of levels from low pollution to levels unlikely to be routinely observed within the U.S. The influences of boundary layer SO 2 on the CIE-weighted UV irradiance are shown in Figure 4. As SO 2 concentration increases, the CIE UV irradiance decreases, although even at very high SO 2 levels the decrease is not substantial. 8.3E-2 8.2E-2 8.1E-2 8.E-2 7.9E-2 7.8E-2 7.7E SO2 concentration, ppb Figure 4. CIE-weighted UV irradiance is plotted as a function of changing SO 2 concentration. The effects of surface ozone levels on UV irradiance are illustrated in Figure 5. As the tropospheric ozone concentration increases, CIE-weighted UV levels at the surface are reduced. The reduction in UV levels is not substantial for typical ozone concentrations, however, and is in fact much less than the effects of changes in cloudiness. 21

22 Q_CIE in W/m2(eff) 8.4E-2 8.3E-2 8.2E-2 8.1E-2 8.E-2 7.9E-2 7.8E-2 7.7E-2 7.6E-2 7.5E-2 7.4E ozone concentration, ppb Figure 5. CIE-weighted UV irradiance is plotted as a function of changing tropospheric (surface) ozone concentration. Cloud and Aerosol Effects Clouds and aerosols are known to have a large impact on UV radiation, reducing the amount of UV reaching the ground by as much as 9%. For this study, we do not distinguish between clouds and aerosols, but use as a definition the amount of attenuation to UVA, specifically 34 nm, at 6 degrees solar zenith angle. This wavelength is often used as a reference to clouds because of its insensitivity to both ozone and sulfur dioxide. The absolute value of UV at 34 nm for 6 degrees solar zenith angle is affected by both elevation and by clouds/aerosols. For an individual site, the elevation, of course, is constant. This feature allows the amount of UV at 34 nm to approximate clouds and aerosols in a manner consistent for all times of year. For most days and most locations, UV is measured twice per day at 6 degrees solar zenith angle. These two values are averaged to give a representation of cloud/aerosol level throughout the day. It is likely that clouds will change throughout any single day. There may even be systematic biases in sampling the cloud indicator in this way, but the method provides a direct procedure to identify attenuation due to cloudiness at an individual site. Ozone Estimates of total column ozone for each location are produced from the spectral UV radiation measurements at each location. The ratio of radiation at 313nm to radiation at 33 nm, often referred to as the A-Pair ratio, gives a value that is roughly proportional to total column ozone. The values at 313 nm and 33 nm are derived from the measurements taken each day at 6 degrees solar zenith angle. For most locations, this 22

23 solar zenith angle selection means that ozone is derived from two scans each day, with the ratio taken from the average radiation at 313 nm and the average radiation at 33 nm. Taking the estimate of ozone directly from the Brewer measurements yields estimates that include both the local tropospheric and directly overhead stratospheric ozone levels. The averaging of the morning and afternoon measurements helps to smooth the effect of tropospheric ozone, which is generally highest in the afternoon. The smoothing will likely help provide an estimate of ozone that is representative of the daytime average level. Partitioning the Effects Over 4, daily values are available for evaluation from the 21 sites. Based on quality control and screening, only 35,811 of those days are used in these analyses. For each site, the quality assured data are analyzed with a separate measure of minimum solar zenith angle, ozone, and cloud cover. Daytime-integrated CIE-weighted UV estimates are used for each location, and regression analysis is performed to determine how much of the variation in UV observed at each location is statistically attributable to each of the three main explanatory parameters: sun angle, clouds and ozone. The statistical formula used to define this relationship for each location is: CIE UV = α cos (SZA) + α clouds + γ ozone + noise (1), where CIE UV is the daytime-integrated CIE-weighted UV value for each day; SZA is the minimum solar zenith angle for each day reported by the Brewer; clouds is the proxy of the radiation measured at 6 degrees solar zenith angle at 34 nm; and ozone is the ration of 313 to 33 nm as described above. The noise term is modeled as independent white noise this approximation is acceptable only because time derivatives are not being derived in the regression equation. To isolate the effects of individual parameters, equation (1) was altered to omit some of the terms. The primary relationships explored were as follows: to identify the percent variability explained by the sun angle: CIE UV = α cos (SZA) + noise (2); to identify the percent variability explained by both the sun angle and ozone variations: CIE UV = α cos (SZA) + γ ozone + noise (3); 23

24 to identify the percent variability explained by both the sun angle and cloud/aerosol variations: CIE UV = α cos (SZA) + α clouds + noise (4); to explain the total percent variability explained by sun angle, clouds/aerosols and ozone variations: CIE UV = α cos (SZA) + α clouds + γ ozone + noise (5). For each parameter, we derive the percent explained variability by calculating the variability described by the statistical formula and dividing by the total variability calculated from each time series for each site. The results are then presented for each station in terms of the percent variability described by each parameter. The results of this analysis of variance are presented in the following section. 24

25 3. Results for Each of the 21 Brewer Locations The results below summarize the percent variability in the day-to-day UV values that can be explained by solar zenith angle, clouds/aerosol, and total column ozone at each location. The first bar in each figure denotes the percent variability explained by solar zenith angle, which has the dominant effect. The second bar denotes the percent variability explained by solar zenith angle and ozone in combination, and the third bar represents the percent explained using solar zenith angle and clouds/aerosols in combination. The methods for deriving the total column ozone and clouds/aerosol proxies used to obtain these results are described in the previous section. The last bar in each figure represents the percent variability in UV explained when solar zenith angle, clouds/aerosol, and ozone are analyzed in combination. The results are presented for each of the 21 EPA Brewer sites. 25

26 Acadia National Park Latitude: 44.4ºN Longitude: 68.3ºW Elevation: 137 m % Variance Explained SZA SZA + O3 SZA + Cld SZA, Cld + O3 Acadia Acadia National Park is the most northeasterly monitoring location in the EPA Brewer network. The figure above demonstrates the relative influences of (i) solar zenith angle (SZA), (ii) SZA and ozone, (iii) SZA and clouds/aerosol, and (iv) SZA, clouds/aerosol, and ozone together in explaining the observed day-to-day UV variability at the site. For Acadia, SZA explains only about half of the observed variability, and adding ozone increases the percent variance explained only slightly. Adding clouds/aerosol to SZA explains over 75 percent of the day-to-day variability at this site, and SZA, clouds/aerosols, and ozone together explain nearly 8 percent of the day-to-day UV variability. 26

27 Albuquerque, NM Latitude: 35.1ºN Longitude: 16.6ºW Elevation: 1615 m % Variance Explained SZA SZA + O3 SZA + Cld SZA, Cld + O3 Albuquerque The metropolitan area of Albuquerque is home to more than 65, people and is known for its dry, hot climate in the summer and generally mild winters. At this location, SZA alone explains more than 8 percent of the observed day-to-day variability in UV, as illustrated above. Adding ozone does not substantially increase the percent variance explained, and adding clouds/aerosol to SZA results in only a slightly larger increase in the percent variance that is explained. SZA, clouds/aerosol, and ozone together explain about 9 percent of the UV variability at the Albuquerque site. 27

28 Atlanta, GA Latitude: 33.8ºN Longitude: 84.4ºW Elevation: 91 m % Variance Explained SZA SZA + O3 SZA + Cld SZA, Cld + O3 Atlanta The twenty-county Atlanta metropolitan area is home to more than four million people. Solar zenith angle (SZA) at this location explains over 65 percent of the observed day-today variability in UV. Adding ozone does not increase this percentage by any nonnegligible amount. SZA and clouds/aerosol together can improve the percent variance explained to over 8 percent. By including all factors SZA, clouds/aerosol, and ozone in the analysis, we can explain approximately 85 percent of the day-to-day UV variability at the Atlanta Brewer site. 28

29 Big Bend National Park Latitude: 29.3ºN Longitude: 13.2ºW Elevation: 329 m % Variance Explained SZA SZA + O3 SZA + Cld SZA, Cld + O3 Bigbend Big Bend National Park covers over 8, acres in south Texas and hosts an array of ecological, paleontological, and archaelogical research. Solar zenith angle (SZA) alone explains over 75 percent of the observed day-to-day UV variability at this more southern monitoring location. Adding ozone does not increase the percentage of variability that can be explained, but adding clouds/aerosol to SZA does substantially improve our ability to account for the variance in UV. SZA, clouds/aerosol, and ozone together explain nearly 9 percent of the observed UV variance at Big Bend National Park. 29

30 Boulder, CO Latitude: 4.1ºN Longitude: 15.2ºW Elevation: 1689 m % Variance Explained SZA SZA + O3 SZA + Cld SZA, Cld + O3 Boulder Boulder s UV monitoring takes place within 3 miles of the city of Denver, a six-county metropolitan area with a population of over 2.4 million people. Solar zenith angle (SZA) explains approximately 7 percent of the observed day-to-day UV variability at the Boulder monitoring location. Adding ozone increases the percent variance explained, though not by a substantial amount. SZA and clouds/aerosol together explain about 8 percent of the UV variance, and including all three factors SZA, clouds/aerosol, and ozone in the analysis increases the percent of the UV variability explained to over 8 percent. 3

31 Canyonlands National Park Latitude: 38.5ºN Longitude: 19.8ºW Elevation: 814 m % Variance Explained SZA SZA + O3 SZA + Cld SZA, Cld + O3 Canyonlands Canyonlands National Park is located in the high desert of Utah. The ecosystem is studied by numerous scientists because of its unique soil, which supports a variety of algae, lichen, and bacteria. Solar zenith angle explains just over 8 percent of the observed day-to-day UV variability at this site. Adding ozone slightly improves the percent variance explained, and SZA and clouds/aerosol together brings the percent explained up an additional, though not large, amount. Together, SZA, clouds/aerosol, and ozone explain approximately 9 percent of the UV variability at Canyonsland National Park. 31

32 Chicago, IL Latitude: 41.8ºN Longitude: 87.6ºW Elevation: 165 m % Variance Explained SZA SZA + O3 SZA + Cld SZA, Cld + O3 Chicago The nine-county metropolitan area of Chicago is home to over eight million people and sits in the heart of the area covered by the EPA Brewer network. Solar zenith angle (SZA) explains slightly more than 6 percent of the observed day-to-day UV variability at this site. Adding ozone increases the percentage explained by only a slight amount, but adding clouds/aerosol to SZA improves the percent variability explained to nearly 8 percent. All three factors SZA, clouds/aerosol, and ozone together explain just over 8 percent of the day-to-day UV variability at the Chicago monitoring site. 32

33 33

34 Denali National Park Latitude: 63.7ºN Longitude: 149.ºW Elevation: 661 m % Variance Explained SZA SZA + O3 SZA + Cld SZA, Cld + O3 Denali Denali National Park is the farthest north site in the EPA Brewer network. The data reflect a higher seasonality than those at any other site, and the winter values are generally very low, often below the detection limit of the instrument. At Denali, the solar zenith angle (SZA) are often very large, as the Sun does not get very high in the sky even during the summer months. SZA at this location explains less than 5 percent of the observed day-to-day variability in UV. Adding ozone does not substantially improve the percent variability explained. SZA and clouds/aerosol together, however, explain about 75 percent of the observed variability. SZA, clouds/aerosol, and ozone in combination explain over 8 percent of the observed variability in the Denali UV measurements. 34

35 Everglades National Park Latitude: 25.4ºN Longitude: 8.7ºW Elevation: 18 m % Variance Explained SZA SZA + O3 SZA + Cld SZA, Cld + O3 Everglades Everglades National Park encompasses more than 1.4 million acres and is one of the more southernly monitoring locations of the Brewer network. At this site, solar zenith angle (SZA) explains over 5 percent of the observed day-to-day UV variability. SZA and ozone explain only slightly more of the observed variance, and SZA and clouds/aerosol together increase the percent variability explained to above 65 percent. Incorporating all three factors SZA, clouds/aerosol, and ozone explains nearly 7 percent of the observed UV variability at Everglades National Park. 35

36 Gaithersburg, MD Latitude: 39.1ºN Longitude: 77.2ºW Elevation: 43 m % Variance Explained SZA SZA + O3 SZA + Cld SZA, Cld + O3 Gaithersburg Gaithersburg, Maryland is located within 3 miles of both Baltimore, Maryland, and Washington, DC, in the eastern U.S. Solar zenith angle (SZA) alone explains just over 6 percent of the observed day-to-day UV variability at the Gaithersburg site. Adding ozone increases the percentage explained only slightly. SZA and clouds/aerosol together explain approximately 8 percent of the variability at the site, and accounting for the influences of SZA, clouds/aerosol, and ozone together explains greater than 8 of the observed day-to-day UV variability. 36

37 Glacier National Park Latitude: 48.7ºN Longitude: 113.4ºW Elevation: 424 m % Variance Explained SZA SZA + O3 SZA + Cld SZA, Cld + O3 Glacier Glacier National Park encompasses over a million acres of land and is located at both a fairly high latitude and high elevation (424 meters). The figure above demonstrates the percent UV variability explained by (i) solar zenith angle (SZA), (ii) SZA and ozone, (iii) SZA and clouds/aerosol, and (iv) SZA, clouds/aerosol, and ozone. SZA alone explains more tha 6 percent of the observed day-to-day UV variability. Adding ozone improves the percent variability to around 7 percent. Adding clouds/aerosol to SZA increases the percent explained to above 75 percent. At this fairly high latitude, high elevation location, SZA, clouds/aerosol, and ozone together explain more than 8 percent of the observed day-to-day variability in the UV measuremets obtained from the Brewer instrument. 37

38 Great Smoky Mountains National Park Latitude: 35.6ºN Longitude: 83.8ºW Elevation: 564 m % Variance Explained SZA SZA + O3 SZA + Cld SZA, Cld + O3 Great Smoky Great Smoky National Park encompasses over a half million acres of land in the eastern United States. At this location, solar zenith angle (SZA) explains less than 7 percent of the observed day-to-day variance in UV. SZA and ozone together explain only slightly more of the variance, but adding clouds and SZA substantially improves the percentage explained. SZA, clouds/aerosol, and ozone in combination explain more than 8 percent of the observed day-to-day variability in UV at Great Smoky National Park. 38

39 Hawaii Volcanoes National Park Latitude: ºN Longitude: ºW Elevation: 1243 m % Variance Explained SZA SZA + O3 SZA + Cld SZA, Cld + O3 Hawaii Hawaii Volcanoes National Park covers over 3, acres. The southerly site is located at a high elevation, making it an interesting place for long-term UV monitoring. Solar zenith angle (SZA) explains less than 4 percent of the observed day-to-day UV variability at this location. SZA and ozone together explain only very slightly more of the variance, while including both SZA and clouds/aerosol boosts the percent explained to over 6 percent. SZA, clouds/aerosol, and ozone in combination explain almost 7 percent of the observed UV variance the Hawaii Volcanoes site. 39

40 Olympic National Park Latitude: 48.1ºN Longitude: 123.4ºW Elevation: 32 m % Variance Explained SZA SZA + O3 SZA + Cld SZA, Cld + O3 Olympic Olympic National Park encompasses over 9, acres and is one of the more northernly parks in the network. Solar zenith angle (SZA) explains around 65 percent of the variance in the day-to-day UV measurements at the park. Ozone and SZA together explain about 7 percent of the observed variance, while clouds/aerosol and SZA explain greater than 75 percent of the UV variability. Incorporating all three factors SZA, clouds/aerosol, and ozone into the analyis improves the percentage explained to greater than 8 percent. 4

41 Research Triangle Park Latitude: 35.9ºN Longitude: 78.9ºW Elevation: 14 m % Variance Explained SZA SZA + O3 SZA + Cld SZA, Cld + O3 R.T.P Research Triangle Park is one of the country s premier research areas located in the eastern U.S. One of the first Brewer instruments was installed near RTP and the data from this site have been studied by a number of researchers. Our analysis of the UV data and the factors influencing its variability indicates that solar zenith angle (SZA) accounts for nearly 7 percent of the observed day-to-day UV variability at this site. Adding ozone to the analysis improves this percentage only slightly, but adding clouds/aerosol to SZA brings the percent explained to over 8 percent. SZA, clouds/aerosol, and ozone in combination explain over 85 percent of the observed UV variance at Research Triangle Park. 41

42 42

43 Riverside, CA Latitude: 34.ºN Longitude: 117.3ºW Elevation: 84 m % Variance Explained SZA SZA + O3 SZA + Cld SZA, Cld + O3 Riverside The Riverside-San Bernardino area near the 9-million person population area of Los Angeles. The location is known both for its air pollution and its success in cleaning its air over the last decade. The figure above illustrates the percent of the day-to-day UV variability explained by (i) solar zenith angle (SZA), (ii) SZA and ozone, (iii) SZA and clouds/aerosol, and (iv) SZA, clouds/aerosol, and ozone together. Including the three factors explains over 9 percent of the variability in the data at this site. 43

44 Rocky Mountain National Park Latitude: 4.ºN Longitude: 15.5ºW Elevation: 2896 m % Variance Explained SZA SZA + O3 SZA + Cld SZA, Cld + O3 Rocky Mountain Rocky Mountain National Park sits on the east end of the Rocky Mountains where its high elevation results in a very strong seasonal cycle. The percent of the day-to-day UV variability explained by the various factors is illustrated above. Solar zenith angle (SZA) accounts for 6 percent of the day-to-day variability. Adding ozone explains a bit more, and adding clouds/aerosol and SZA explains over 75 percent of the variance. Adding SZA, clouds/aerosol, and ozone together explains nearly 8 percent of the UV variability at this site. 44

45 Sequoia National Park Latitude: 36.5ºN Longitude: 118.8ºW Elevation: 549 m % Variance Explained SZA SZA + O3 SZA + Cld SZA, Cld + O3 Sequoia Sequoia and King s Canyon National Parks are over 8, acres in size and host roughly a million visitors each year. For this site, much (greater than 75 percent) of the day-to-day UV variability is explained by solar zenith angle (SZA) alone. Adding ozone explains slightly more, and adding by clouds and SZA together we can explain a few more percent. When SZA, clouds/aerosol, and ozone are all taken into account, they explain approximately 9 percent of the UV variability at this site. 45

46 Shenandoah National Park Latitude: 38.5ºN Longitude: 78.4ºW Elevation: 325 m % Variance Explained SZA SZA + O3 SZA + Cld SZA, Cld + O3 Shenandoah Shenandoah National Park covers roughly 2, acres and hosts over a million visitors each year. The relative influences of (i) solar zenith angle (SZA), (ii) SZA and ozone, (iii) SZA and clouds/aerosol, and (iv) SZA, clouds/aerosol, and ozone on the day-to-day UV variability are shown above. SZA alone explains over 6 percent of the day-to-day UV variability at this site. Adding ozone does not improve this percentage, but including clouds/aerosol increases the percent variance explained to over 8 percent. 46

47 Theodore Roosevelt National Park Latitude: 46.9ºN Longitude: 13.4ºW Elevation: 238 m % Variance Explained SZA SZA + O3 SZA + Cld SZA, Cld + O3 T. Roosevelt Theodore Roosevelt National Park, with 7, acres, hosts more than a half million visitors each year. The figure above illustrates the ability of (i) solar zenith angle (SZA), (ii) SZA and ozone, (iii) SZA and clouds/aerosol, and (iv) SZA, clouds/aerosol, and ozone to explain the observed day-to-day variability in UV. SZA alone explains over 6 percent of the UV variability. Adding ozone improves this percentage slightly, but the more substantial improvement comes when clouds/aerosol are added. SZA, clouds/aerosol, and ozone together more than 8 percent of the UV variability at this site. 47

48 48

49 Virgin Islands National Park Latitude: 18.3ºN Longitude: 64.8ºW Elevation: 3 m % Variance Explained SZA SZA + O3 SZA + Cld SZA, Cld + O3 Virgin Islands Virgin Islands National Park is the most southern location in the EPA network and is located within the tropics. At this site, the percent of day-to-day UV variability explained by solar zenith angle (SZA) is smaller than has been typically observed for the midlatitude sites. Adding ozone does not increase this percentage, as ozone does not vary much throughout the year at this latitude. Adding clouds/aerosol to SZA improves the percent variance explained to over 65 percent, and including SZA, clouds/aerosol, and ozone together explains nearly 7 percent of the day-to-day UV variability. 49

50 4. Results for the Entire Brewer Network The analyses presented in this section compare the relative influences of solar zenith angle (SZA), clouds/aerosol, and ozone on UV variability across the entire EPA Brewer network. The changes in SZA and ozone over time of year for each of the sites have already been documented. Clouds/aerosol behave in a more random fashion and changes can occur over very small time and spatial scales, so their influences on UV are likely to not follow any consistent geographic or other pattern. Table 2 summarizes how the sites compare for the portion of daily variability that can be explained statistically by the solar zenith angle, ozone, and clouds/aerosol. The second and third columns list the latitude and elevation for each site. The fourth column (yellow) lists the percentage of variance in CIE daytime-integrated UV levels explained by the minimum solar zenith angle observed at noontime for each day. The fifth column (light purple) gives the percentage of explained variance when both solar zenith angle and ozone are included as explanatory variables. The sixth column gives the percentage of explained variance when solar zenith angle and clouds (gray) are used. The seventh column (blue) gives the percentage of explained variance when solar zenith angle, clouds, and ozone are used as explanatory variables. The final column gives the standard deviation in the daily integrated CIE weighted UV levels. The results are, as is always the case, dependent on the proxies used for cloud and ozone estimates and on the statistical model used. Other independent methods of estimating these parameters may have other benefits. However, the direct methods for estimating clouds using transmission at 34 nm (6 degrees solar zenith angle) and estimating ozone from the A- pair ratio (313 nm to 33 nm) allows for the minimization of many sampling and colocation issues when external datasets are used. Because these issues are minimized, many fundamental insights can be gained from the analysis. 5