Obtaining a Better Understanding of Ozone Air Pollution in the Four Corners Area via Comprehensive Analyses of Ambient Air Monitoring Data Control # 543 DRAFT MANUSCRIPT SUBMITTED 1/7/05 Mark E. Sather U.S. Environmental Protection Agency (EPA), Region 6, Multimedia Planning and Permitting Division, Air Quality Analysis Section (6PD-Q), 1445 Ross Avenue, Dallas, TX 75202 ABSTRACT During the summer of 1997 the New Mexico Environment Department (NMED) deployed an ambient ozone sensor near Farmington, New Mexico, in the northwest corner of the State. Over a short time the 8-hour ozone design value at this site rose to become the highest ozone concentration sensor in the Four Corners area encompassing northwestern New Mexico, southwestern Colorado, southeastern Utah, and northeastern Arizona. The quick rise in northwestern New Mexico ozone concentrations prompted immediate attention from citizens and State, Local, Federal, and Tribal governments, and additional ambient monitoring of ozone and ozone precursors has occurred over the past few years in New Mexico, Colorado, and Arizona. This paper will provide a comprehensive analysis of ambient air monitoring data collected in the Four Corners area and adjacent region since the early 1990’s. Trends, high ozone episode, and Four Corners area/regional ozone background analyses will be conducted using the ozone, nitrogen oxides (NOx), and meteorological data recorded by all of the ambient monitors in the Four Corners area and adjacent region. In addition, results from a passive ozone monitoring saturation study completed in 2003 in San Juan County, New Mexico, will be discussed, and the initial results from first time long-term volatile organic compound (VOC) canister sampling at two sites in San Juan County, New Mexico, in 2004 will be presented. All of these data assessments will help obtain a better understanding of the current ozone air quality in the Four Corners area. INTRODUCTION Four Corners area ambient ozone measurements began being recorded at an adequate percentage of data capture in 1994 at two sites in southwestern Colorado: (1) Mesa Verde National Park and (2) La Plata County. In 1997 the NMED deployed their first ambient ozone sensor about 20 miles west of Farmington at Substation. Later that year a second ozone monitor was added in La Plata County, and finally in 2000 the NMED installed another ozone monitoring site about 15 miles east of Farmington at Bloomfield. All of these monitoring sites, plus regional ozone monitoring sites at Canyonlands National Park in Utah, and at Grand Canyon National Park and Petrified Forest in Arizona, are displayed in Figure 1. The major NOx stationary source emissions (NOx emissions > 100 tons per year) are also shown in Figure 1 with San Juan County, New Mexico containing the greatest amount of stationary source NOx emissions in the Four Corners area. 1 Figure 1: Four Corners Area and Regional Ozone Monitors and Major NOx Stationary Emission Sources. 2 Within a short time the ozone concentrations at the Substation monitor rose to become the highest ozone concentration sensor in the Four Corners area. The 8-hour ozone concentration trend lines are depicted in Figure 2 for all Four Corners area monitors plus regional ozone monitoring sites at Canyonlands and Grand Canyon for greater perspective. The 8-hour ozone National Ambient Air Quality Standard (NAAQS) is 0.08 ppm with a maximum allowable value of 0.084 ppm (84 ppb)1, and the 8-hour ozone design value is defined as the three year average of the annual fourth highest values.2 The first complete year of ozone data at the Petrified Forest site was 2003, thus not enough data has been collected for an initial 8-hour design value. When ozone monitoring began at Substation its initial 8-hour ozone design value of 67 ppb was just above Mesa Verde and below all of the other area and regional ozone monitoring sites. However, by the three year period 2000-2002, the Substation 8-hour ozone design value (and the Bloomfield initial 8-hour ozone design value) had reached 76 ppb and was above all of the other area and regional ozone monitoring sites. An additional interesting insight from studying the trend lines is that beginning in the period 1999-2001 the two La Plata County trend lines go sharply down in opposition to the other ozone monitoring sites. Personal communication with U.S. EPA Region 8 air monitoring personnel indicate localized NOx scavenging may be occurring at these sites from the addition of many oil and gas wells very close to the La Plata County ozone monitors. The two ozone monitors track very well indicating that this is probably not a mechanical problem with the La Plata ozone samplers. Figure 2: Ozone 8-hour Concentration Trends in the Four Corners Area and Outlying Region. 3 ANALYTICAL METHODS The ozone, NOx, and meteorological data for all sites were either downloaded from the U.S. EPA Air Quality System (AQS) national database or received from the NMED. Passive ozone data were received from the U.S. EPA Region 6 Houston Laboratory after collection by the NMED and many San Juan County, New Mexico citizens, and from the University of New Mexico (UNM) School of Medicine. Software packages used to analyze the data included SAS, Environ’s Ambient Monitoring Data Analysis System (AMDAS), Excel and Lotus 1-2-3 spreadsheet, Lotus Freelance Graphics, and Geographic Information Systems (GIS). RESULTS AND DISCUSSION Four Corners Area Ozone and Meteorological Data Analyses When the 8-hour ozone design value at the Substation site went quickly from 67 ppb to 76 ppb, effectively cutting the site’s NAAQS “cushion” in half, it got the attention of citizens and State, Local, Federal, and Tribal officials. One of the first questions asked was “Is the Substation 8hour ozone concentration trend different or similar to other area and regional ozone monitoring sites?” As seen in Figure 2, the Substation ozone concentrations rose at a higher rate than all the other sites, including the nearby rural Mesa Verde site. All Four Corners area ozone monitors, including passive ozone monitors which will be discussed later in this paper, are shown in Figure 1. All sites sit at high elevation (from 1532-2165 meters above sea level) amongst complex arid mountainous terrain. The Mesa Verde site is the highest elevation site at 2165 meters above sea level. This site, about 30 miles north of Substation, can be considered to represent the top bound of total background ozone for the Four Corners area. Total background ozone is defined in this paper, as similarly discussed by Fiore et. al.3, as natural background ozone concentrations from the stratosphere and from photochemical reactions of ozone precursors from natural sources such as lightning, wildfires, soils, and vegetation, plus anthropogenic background ozone concentrations from global hemispheric transport and from regional transport. The Mesa Verde ozone data is considered a top bound of total background ozone estimates for the Substation and Bloomfield monitors in San Juan County, New Mexico because of its higher elevation (and thus a greater amount of background ozone concentrations from the stratosphere) and because some of the regional transported ozone concentrations at Mesa Verde come from local ozone production around the Substation and Bloomfield monitors. On average, about 25% of the time the resultant wind directions recorded at Mesa Verde in the summer come from the SE sector which contains the area around the Substation and Bloomfield monitors. Ozone Diurnal Profile Analysis When reviewing the ozone diurnal profiles for the Four Corners area continuous ozone monitors (Figure 3) the Mesa Verde site stands out as the only trace exhibiting a generally flat diurnal profile, a desired characteristic for a good rural ozone background site. The Substation, Bloomfield and La Plata County sites all have their ozone concentrations decrease more at night and early in the morning in response to impact from local anthropogenic NOx emissions. The Mesa Verde site is negligibly affected from anthropogenic locally generated ozone and ozone precursors, having its ozone concentrations primarily come from natural sources (including the stratosphere), hemispheric transport, and regional transport. The Mesa Verde ozone diurnal 4 profile showed a small increase from 1998 to 2003. Note how the Substation site ozone diurnal profile was just a little above the Mesa Verde profile in 1998, but in 2003 the afternoon difference between the profiles was much more noticeable at a maximum of 6 ppb (the result of increased local ozone production at the Substation site). Similarly, the Bloomfield site 2003 ozone diurnal profile also exceeded the Mesa Verde profile at a maximum of 7 ppb. In contrast to the notable ozone concentration increase at the Substation site the La Plata sites’ ozone traces decreased significantly from 1998 to 2003, possibly due to localized NOx scavenging as surmised by U.S. EPA Region 8 air monitoring personnel. Figure 3: Four Corners Area Ozone Diurnal Profiles. 5 High Ozone Episode Day Analysis There have been two 8-hour ozone exceedance days in the Four Corners area, both in August, 2000, and both at the Bloomfield site in San Juan County, New Mexico. The exceedances were just one ppb over the standard at 85 ppb. The ozone diurnal profiles in Figure 4 show high background concentrations on both days, inferred from the Mesa Verde trace. Assuming the peak Mesa Verde 8-hour ozone concentration of 82 ppb on August 2, 2000, was total background for the Four Corners area, it took around 3 ppb (8-hour average) of local ozone production to put Bloomfield over the standard at 85 ppb. Similarly on August 4, 2000, around 5 ppb of local ozone production (8-hour average) was added on to the assumed total background of around 80 ppb (from the Mesa Verde site) to put Bloomfield over the standard at 85 ppb. Remember that the 3-5 ppb (8-hour average) of local ozone production are low estimates because the local amount of ozone production around Bloomfield that gets transported to the NW is included in the Mesa Verde total ozone background concentration. An additional key nugget of information derived from the August 2, 2000, exceedance day diurnal profile plot is the difference between the Four Corners area and Houston Baytown traces. Note how the Houston Baytown trace jumps suddenly 49 ppb from hour 11 to hour 12. This jump is indicative of ozone forming from chemical reactions between NOx and highly reactive volatile organic compounds (HRVOC). The August 2, 2000, and August 4, 2000, Four Corners area profiles do not show this large hourly jump in ozone concentration, at most rising by 18 ppb in one hour at the Bloomfield site on August 4, 2000. Thus, on the two 8-hour ozone exceedance days in August of 2000 in the Four Corners area the high ozone was most likely formed by photochemical reactions between NOx and more slowly reacting VOC such as propane and nbutane. The propane and n-butane concentrations would be largely coming from the oil/gas industrial activity in the Four Corners area, an industry that has significantly expanded in the early 2000’s. Limited (i.e. 30 second grab samples during one week) VOC canister monitoring in 2002 showed elevated levels of light alkanes above background levels in the Four Corners area.4 Ozone Concentration Trends Analysis from 1997-2003 To determine how much of the ozone concentration increase at Substation from 1997 to 2003 was due to local anthropogenic sources, peak afternoon 8-hour ozone concentrations were analyzed at all Four Corners area sites, at the two regional ozone sites at Grand Canyon and Canyonlands, and at the high ozone site in the city of Albuquerque. In order to smooth out the year to year effects of meteorology the three year period 1997-1999 was compared to the four year period 2000-2003. The Mesa Verde site represents a top estimate for background changes in the Four Corners area, and similarly, the Grand Canyon and Canyonlands sites give a greater perspective by representing regional background changes to the west of the Four Corners area. Results of the afternoon 8-hour ozone concentration analyses are shown in Table 1. The Mesa Verde site recorded a 10% increase in afternoon 8-hour ozone concentrations from 1997-1999 to 2000-2003, a little above the regional background increase to the west. The City of Albuquerque recorded a little larger increase of 12% while the two La Plata sites showed a 12% decrease, possibly due to the samplers being affected by new close by NOx sources. The Substation site 6 Figure 4: Four Corners Area Ozone Diurnal Profiles on Two Exceedance Days. 7 recorded the largest increase at 17%, 7% above the Four Corners background site at Mesa Verde. Thus, taking into account that some of the 10% ozone increase at Mesa Verde was due to local ozone production around Substation, roughly half of Substation’s 17% increase in 8-hour ozone concentrations was due to local anthropogenic ozone production and the remaining half due to an increase in the Four Corners area background ozone concentrations. Helping produce the higher ozone concentrations were the higher ambient temperatures recorded from 2000-2003. In Figure 5 note the strong correlation (r = 0.94) between mean temperature and mean afternoon 8-hour ozone concentrations at the Substation site. The Mesa Verde site also had a high correlation coefficient of 0.80. Stagnation, indicated by the hourly wind speed profile, did not increase appreciably during the period 2000-2003, but there was an increase in NW winds during this same period as compared to 1997-1999. This wind direction shift contributed to the higher ozone concentrations recorded from 2000-2003 at Substation by keeping ozone-laden air around longer in the Substation area through flow reversal. Figure 6 shows an increase in afternoon NW winds in 2000 compared to 1999, and this same wind directional pattern occurred from 20012003. Table 1: Comparison of 1997-1999 to 2000-2003 Afternoon 8-hour Ozone Concentrations; June-August (0800-1500, 0900-1600, 1000-1700, 1100-1800 LST) Site % Change Substation 17 Albuquerque 12 Mesa Verde 10 Grand Canyon 8 Canyonlands 7 La Plata 7001 -12 La Plata 7003 -12 8 Figure 5: Substation Site Ozone and Meteorological Data Trends. Hourly Av erage Wind Speed Profile (m ph) 100 0-4 % o f to ta l 80 5-9 >= 10 60 40 20 0 97 98 99 00 01 02 03 Ye a r Hourly Av erage Wind Direc tions by Sec tor 100 NE % o f to ta l 80 SE SW NW 60 40 20 0 97 98 99 00 01 02 03 Ye a r 85 70 80 60 ppb d eg . F M ean Tem perature and 8-hour O z one 75 50 Mean T Mean 8-hr O3 70 40 97 98 99 00 01 02 03 Ye a r R2=0.89 (r=0.94) Mean 8-hr O3 computed from 8-hour averages ending at hours 1500, 1600, 1700, and 1800 LST Mean T computed from hourly averages from 0500-1800 LST 9 Figure 6: Substation Wind Sector Analysis for 1999 and 2000. Four Corners NOx Data Analyses The detailed ozone and meteorological analyses presented in the previous section showed both total background and local anthropogenic components of the increased 8-hour ozone concentrations recorded at the Substation site. The local anthropogenic component was formed by photochemical reactions of the NOx and VOC emissions output by local industry and mobile sources. VOC data has not been collected in the Four Corners area until recently but NOx data has been collected at the Substation and Bloomfield sites since 1997, and the La Plata sites began collecting NOx data in 2001. The 2003 NOx diurnal profiles in Figure 7 show the highest NOx concentrations at the Bloomfield and Substation sites. Looking at the profile reveals the following: (1) a strong 4-7 AM LST peak at the Bloomfield site indicative of mobile source morning commute emissions; (2) a little lower peak in NOx concentrations at the Substation site two hours later as the wind shifts to the SE bringing Bloomfield and Farmington aged NOx emissions to the site along with the fresh NOx emissions from the nearby San Juan and Four Corners Power Plants; and (3) the peak NOx concentrations at Bloomfield and Substation are similar to the peak morning NOx concentrations recorded at larger city sites such as Baton Rouge, Louisiana. Looking at morning NOx concentration trends reveals increases since 1997, 10 especially at the Bloomfield site. This increase was not due to increased stagnation or major wind direction changes as seen in the Bloomfield trends in Figure 8. Tracking the ambient NOx concentrations in the Four Corners area will continue to be critical in assessing the effects of future industrial growth. Currently there are three new power plant applications under review in the Four Corners area.5 Figure 7: Mean NOx Diurnal Profiles. 11 Figure 8: Bloomfield NOx and Meteorological Data Trends Hourly Av erage Wind Speed Profile (m ph) 100 0-4 % o f to ta l 80 5-9 >= 10 60 40 20 0 97 98 99 00 01 02 03 Ye a r Hourly Res ultant Wind Direc tions by Sec tor 100 NE % o f to ta l 80 SE SW NW 60 40 20 0 97 98 99 00 01 02 03 02 03 Ye a r NO x Conc entrations 40 ppb 30 20 10 0 97 98 99 00 01 Ye a r 12 San Juan County, New Mexico Passive Ozone Monitoring Saturation Study In 2003 a passive ozone monitoring saturation study was conducted in San Juan County, New Mexico by the U.S. EPA Region 6 and NMED in order to obtain a better picture of the distribution of high ozone concentrations in the County beyond the two existing continuous ozone monitoring stations at Substation and Bloomfield, both located in the north central part of the County. Passive ozone monitors were deployed because of their low cost and simplicity of use. Seven passive ozone sites (monitors #6-#12 shown in Figure 1) were set up to supplement the two existing continuous ozone sites. The passive sites were located in a variety of areas including more densely populated areas, downwind rural areas, and different areas of complex terrain such as on mesas/mountains or in valleys. The passive ozone saturation study ran for six weeks from July 21-August 29, 2003. Monitoring results, shown in Figure 9, showed significantly high ozone concentrations in the western and northeastern areas of San Juan County, New Mexico, in addition to the high ozone concentrations already found in the north central area of the County. These results have helped the NMED in selecting an additional continuous ozone monitoring site for the Four Corners area. With the financial assistance of the Bureau of Land Management, the NMED is considering locating a new continuous ozone monitor in western Rio Arriba County, close to the area that recorded high passive ozone concentrations in the 2003 saturation study. In addition to the U.S. EPA Region 6/NMED passive ozone monitoring saturation study, from June-October, 2003, the Community Environmental Health Program at the UNM School of Medicine collected weekly passive ozone samples at four sites (shown in Figure 1 as the two squares in La Plata County and the two squares in San Juan County, New Mexico), plus a collocated site at Bloomfield. The U.S. EPA Region 6 analyzed the UNM data that overlapped with the saturation study (i.e. the five weeks from July 23-August 27, 2003). Results of the two overlapped studies nicely complimented each other. San Juan County, New Mexico VOC Baseline Monitoring After the 2003 passive ozone monitoring saturation study and review of the historical ozone, NOx, and meteorological data, the U.S. EPA Region 6 recommended that whole summer season VOC monitoring data be collected for the first time in the Four Corners area. This would set a current baseline of VOC concentrations in the Four Corners area based upon current meteorological conditions and emissions outputs. This VOC baseline could then be compared to other areas with ozone pollution problems such as Baton Rouge, Louisiana, and help in checking current VOC emission inventory estimates. The baseline data can also be compared to future ambient data collected in order to corroborate the implementation of any future VOC emission controls. Under current review are plans for drilling 10,000 gas wells over the next 20 years in the Four Corners area.5 Thus, in July 2004, the U.S. EPA Region 6 and the NMED began collecting VOC canister data at two sites in San Juan County: (1) Bloomfield and (2) Navajo Lake State Park close to passive ozone site #10. Data was collected from July–September on the national 1/6 day VOC 24-hour canister schedule, and data will also be collected May–June, 2005, also on the national 1/6 day schedule. Three-hour and 24-hour samples were collected at the Bloomfield site. The three-hour sample was collected from 4-7 AM LST when the site is known to measure the highest NOx concentrations. Three-hour samples only were taken at Navajo Lake 13 State Park from 3-6 PM LST in order to focus on the contributions from biogenic sources. As of the time of drafting this paper the VOC data were still being processed and analyzed by the U.S. EPA Region 6 Laboratory in Houston. Once the final results are received they will be reported in this section of the paper. Figure 9: San Juan County Passive Ozone Monitoring Study Results. 14 CONCLUSIONS This paper presented the results of a comprehensive analysis of ambient monitoring data in the Four Corners area and adjacent region. This data assessment helped obtain the following understanding of ozone pollution in the Four Corners area: 1. Ozone concentrations at the Substation monitoring site in northcentral San Juan County, New Mexico rose from 1997-2003 to become the highest ozone concentration sensor in the Four Corners area. 2. Total background ozone concentrations in the Four Corners area, as represented by the Mesa Verde site, did rise from 1997-2003, but the Substation site ozone concentrations rose even higher than Mesa Verde and at a higher rate than the City of Albuquerque. On the two 8-hour ozone exceedance days in August, 2000, at the Bloomfield site, local ozone produced (8-hour average) was estimated to be at least 3-5 ppb above total background ozone concentrations. At least 50% of the afternoon 8-hour ozone concentration increase at the Substation site from 20002003 compared to 1997-1999 was due to local anthropogenic sources, and the remaining 50% due to a rise in total background ozone concentrations. Total background ozone concentrations from the Mesa Verde site are considered to be a top estimate for the Four Corners area because of the high elevation of the site and because 25% of the time during the summer resultant wind directions at Mesa Verde come from the SE sector which contains the area around the Substation and Bloomfield monitors. 3. Ozone concentrations at Substation and Mesa Verde correlate well with ambient temperature, and temperatures and ozone concentrations were higher during the period 2000-2003 as compared to 1997-1999. In addition, afternoon wind directions from 2000-2003 were more favorable to keeping high ozone concentrations in the Substation area as compared to 1997-1999. 4. NOx emissions are most plentiful in San Juan County, New Mexico, and ambient NOx concentrations rose at both the Substation and Bloomfield sites during the period 2000-2003 as compared to 1997-1999, especially at Bloomfield which is more impacted by mobile source NOx emissions. The wind speed and wind direction data were comparable during each comparison period which indicates that the NOx concentration increase came from increased anthropogenic (mobile source and stationary source) activity. 5. The 2003 passive ozone monitoring saturation study in San Juan County, New Mexico showed significant ozone concentrations in the western and northeastern part of San Juan County in addition to the high ozone concentrations already found in the northcentral part of the County. The NMED is using this data to help site an additional ozone monitor in western Rio Arriba County. 6. First time long-term VOC monitoring data was collected in 2004 to set a current VOC concentration baseline for the area, and to compare the VOC concentrations to other areas with ozone pollution problems. Future VOC monitoring can be compared to this baseline to determine a trend and to help document the “before and after” impact of any future VOC emission reductions. 15 In conclusion, this paper presented many analyses of ambient monitoring data to better understand the current 8-hour ozone concentration trends in the Four Corners area. The U.S. EPA Region 6 is encouraged by the interest of citizens and the State/Local/Tribal and other Federal government representatives to work right now to make sure ozone air pollution in the Four Corners area does not deteriorate in the future where it would get close to or exceed the public health/welfare 8-hour ozone standard. ACKNOWLEDGMENTS The author is very grateful to many people for contributing to this project: Tim Booker, Joe Cotie, Josephine Ball, Andy Berger, Roger Polisar, Erik Aaboe, Mary Uhl and Sandra Ely of the NMED, Ed 0’Neil of the U.S. EPA Region 6 Houston Laboratory, and Melody Lister and Jeff Danielson of Lockheed Martin for preparation of the excellent GIS maps used in this paper and the oral presentation. REFERENCES 1. 40 CFR Part 50.10 – National 8-hour primary and secondary ambient air quality standards for ozone, July 1999 ed., pg. 9. 2. 40 CFR Part 50 – National Primary and Secondary Ambient Air Quality Standards, Appendix I – Interpretation of the 8-hour Primary and Secondary National Ambient Air Quality Standards for Ozone, July 1999 ed.; pg. 65. 3. Fiore, A.; Jacob, D.J.; Liu, H.; Yantosca, R.M.; Fairlie, T.D.; Li, Q. J. Geophys Res. 2003, 108, 4787-4798. 4. Katzenstein, A. S.; Doezema, L.A.; Simpson, I. J.; Blake, D.R.; Rowland, F.S. PNAS. 2003, 100, 11975-11979. 5. “Utilities Apply to Construct Power Plants Near Parks”; Washington Post, Friday Oct. 29, 2004; pg. A03. KEY WORDS Four Corners Area, Ozone, Trends, Nitrogen Oxides, Volatile Organic Compounds 16