Atmospheric Chemistry Measurements at Whiteface Mountain, NY: Ozone and Reactive Trace Gases

Measurements of ozone and reactive trace gases spanning four decades at the Whiteface Mountain summit observatory are presented. Ozone (O3) measurements began in the mid-1970’s, and acid rain and O3 precursor gas measurements became routine in the late 1980’s and early 1990’s. Measurements at the lower altitude lodge level have also been performed routinely since about 2000. The 40-year O3 record shows up and down fluctuations through the 1980’s, a relatively stable period into the early 2000’s, and indications of a decreasing trend over the past ten years. Sulfur dioxide (SO2) and carbon monoxide (CO) trends are clearly decreasing over the roughly 25-year period of measurements at the summit observatory. Oxides of nitrogen (NOy and NO2) show rather more complicated trends, increasing to a maximum in the mid-2000’s, and decreasing sharply until 2011 with slight increases in concentration since then. Wind rose analysis shows the greatest contribution to high concentrations of precursor gases are from the west, southwest, and southern sectors, with SO2 and oxides of nitrogen having the most sharply defined high pollution sectors. Seasonal variations of trace gas concentrations at the summit and lodge levels are also examined. Ozone concentrations are highest in the spring months at both locations, and higher at the summit than the lodge. In contrast precursor gases (SO2 and NOx) show highest concentrations in winter months with the lodge consistently higher than the summit.


INTRODUCTION
Whiteface Mountain is located in the northern part of the Adirondack State Park in upstate New York, North America.On Whiteface summit (1483 m) is the University of Albany's Atmospheric Sciences Research Center's (ASRC) observatory, where meteorology and atmospheric chemistry monitoring have taken place for over four decades.Trace gas monitoring of ozone (O 3 ) started in 1974 (Mohnen et al., 1977); and measurements of carbon monoxide (CO), sulfur dioxide (SO 2 ), and oxides of nitrogen (NO, NO 2 and NO Y ) began in earnest in the late 1980's to develop an understanding of transport and chemical processing of atmospheric pollutants; and to quantify the efficacy of pollution control policy (Parrish et al., 1993;Trainer et al., 1993).Aerosol, precipitation, and cloud chemistry sampling complement the trace gas monitoring and collectively these observations contribute to our understanding of ozone and PM 2.5 pollution, acid precipitation chemistry and to the deposition processes of atmospheric pollutants (Schwab et al., 2016b).These measurements, and the trend data produced over long measurement periods are important as states and nations continue air pollution reductions to protect human health and the environment, and mitigate climate change (these pollutants variously act as absorbers or scatterers of incoming and outgoing radiation, thereby affecting the planet's radiative balance).
As noted in our companion paper (Schwab et al., 2016a) mountain top observatories offer a number of advantages for air pollution measurements.Most importantly, this site is located hundreds of km from any major pollution sources, so it is a good indicator of regional background, air quality and provides an opportunity to check the response to regional emission controls as well as providing measurement data for the validation of air quality modeling results.These characteristics along with the long data record are the main reason for the extensive use of Whiteface Mountain data in numerous publications over the years (Logan, 1989;Lefohn et al., 1990;Aneja and Li, 1992;Cooper et al., 2012;Cooper et al., 2014).Whiteface summit is a lone massif peak, north of the main group of Adirondack high peaks, and typically influenced by aged air masses that predominantly flow from the west and southwest transporting emissions from source regions in the Midwest.Less typical northerly flows are associated with clean air from Canada and the arctic.These two distinct source regions provide a clear contrast between polluted air masses originating in the south and Midwest versus generally clean air from the north -which is an indicator of background pollution levels.In this paper, the Whiteface Mountain sites are presented as locations where trace gas chemistry measurements have been performed for decades and are ongoing.
Ozone trends at remote mountain sites have attracted additional interest in recent years, and this has been accentuated by the recent decision of the USEPA to lower the National Ambient Air Quality Standard for ozone to 70 ppb over an eight hour period (EPA, 2015).Due to the fact that such sites have large fetches and are sometimes or even frequently sampling air from the free troposphere, scientists often use these sites as an indication of the "North American Background" ozone levels, defined as the ozone levels that would be present in the absence of North American anthropogenic emissions (Jaffe et al., 2004;Fiore et al., 2014).This very important parameter has a large impact on the ability of states and metropolitan areas to meet ever more stringent ozone standards.
Over the course of these long term observations trace-gas analyzers and data acquisition systems have been upgraded to take advantage of improved accuracy and updated signal processing and I/O with onboard digital computers.With a general decrease in pollution levels, analyzers with enhanced sensitivity, such as Thermo and Teledyne API trace-level instruments have been implemented.Data acquisition and storage, initially using on-site magnetic tape, have been updated with internet capable dataloggers such as the Campbell CR3000 that upload data files daily; and can be accessed directly via a web interface which greatly improves the response time to problems with the instrumentation and data stream.ASRC operates two environmental monitoring sites at this location.The Marble Mountain Lodge site is located on the eastern shoulder of the Whiteface Massif at 44.393°N and 73.859°W with an elevation of 604 m above sea level.Headquarters, offices and laboratories for the operation are located at the lodge level.The Marble Lodge measurement site is situated in a clearing of a wooded area that consists primarily of a mixed Northern Hardwood forest type.Additional details and a map showing the location are in our companion paper (Schwab et al., 2016a).

PARAMETERS MEASURED: INSTRUMENT, SAMPLING, AND METHOD DESCRIPTION
Trace gases monitored at Whiteface include O 3 , CO, SO 2 , NO, NO 2 and NO Y .Tables 1(a) and 1(b) list the gases measured year round and pertinent information for the summit and lodge sites, respectively.A timeline and additional instrument and method details are presented in Tables S1 and S2 (except for ozone, which is addressed below).Other parameters not discussed in this paper are also measured, and many of the condensed phase species are presented and discussed in a companion paper (Schwab et al., 2016b).Thermo Environmental (TEI) and Teledyne API continuous trace gas analyzers have been the instruments of choice due to the reliability and design modifications employed to measure ever lower concentration levels.Even these instruments are severely challenged to measure the quite low concentrations often experienced at both summit and lodge locations.In addition to the gas concentration measurements, basic meteorological parameters such as temperature, relative humidity, wind speed, wind direction, and barometric pressure are measured and recorded at each location.
Two TEI 42CTL analyzers monitor nitrogen oxides at the summit, and both have been modified to 1) improve specificity for NO 2 through use of photolytic conversion to NO; and 2) reduce sampling losses of nitric acid and other "sticky" oxides of nitrogen by relocation of the heated molybdenum converter.For NO 2 conversion, an Air Quality Design UV photolytic converter replaces the standard heated molybdenum converter.For NO Y conversion the heated molybdenum converter has been moved out of the analyzer to within 20 cm of the intake manifold to minimize sample exposure to long inlet lines and valves and tubing within the 42CTL.
At the summit observatory ambient air for the NO x (NO + NO 2 ), SO 2 , and CO analyzers is sampled from a 6 m FEP Teflon 36 mm diameter tube that extends 1.6 m above the observatory's roof with an FEP lined stainless-steel cap to protect from precipitation.A Dayton blower capable of 50 CFM in free air is installed at the downstream end of the sample manifold which is simply four press-fit holes in the 36 mm FEP tube.Exhausts from the Dayton blower, analyzers, and calibration systems are collected in a 10 cm diameter PVC exhaust manifold, then exhausted through a Soler & Palau in line duct fan.The exhaust fan is capable of 293 CFM in free air to ensure wind caused pressure differences will not cause flow reversal.The inlet to the O 3 instrument is a 1/2" diameter PFA tube from the roof about 6 m length which is stepped to ¼" diameter with a PFA fitting and an inline 5 m pore size PTFE Teflon filter feeding the sample port of the analyzer.5 m pore size PTFE Teflon filters are also used on the sample lines for other analyzers to avoid particulate contamination of the analyzer optics.At the lodge level, the sampling manifold is a 5 cm diameter glass manifold maintained by NYSDEC.
Two-point calibration of the trace gas analyzers is done manually on a weekly schedule.DEC staff calibrates and maintains the 49C O 3 analyzers, while ASRC operates the CO, SO 2 NO/NO 2 and NO Y analyzers at the summit and CO and NO x analyzers at the lodge.The 49C O 3 analyzer is checked, audited, and calibrated when required using a DEC certified calibrator, while the other analyzers are calibrated with custom dynamic dilution calibration systems that use Tylan mass flow controllers set with Tylan control boxes, Furon PTFE or Kip stainless steel valves, PTFE mixing chambers and either Scott-Marin primary standard calibration gas cylinders (CO and NO) or VICI certified permeation tube systems (SO 2 and NO 2 ).The SO 2 permeation device is kept at a stable output using a flow and temperature controlled Kin-Tek oven while the NO 2 permeation device is installed in a PTCS-2 oven with an internal Tylan mass flow controller.
Calibration valves are actuated either manually on site or automatically by a Campbell CR3000 data logger.Automatic calibration checks for ASRC operated instruments (see Table 1) occur four times daily at the summit observatory and once daily at Marble lodge.Calibration worksheets, based on pdf forms, are easily compiled to a database to streamline data processing which uses calibration coefficients and calibration checks to correct for analyzer drift and to assign QA flags.Calibrator flow controllers are audited annually and corrections are employed to ensure stable and accurate calibrations.Analysis of raw 1-minute average data collected with the CR3000 provides annual tables of 1-hour average data and QA flags following the format of the North American Research Strategy for Tropospheric Ozone (NARSTO) data archival protocol.Trace gas and meteorological data are available as 1-hour averages with a QA flag for each parameter.The data files are individual annual compilations in NARSTO format and are delimited with commas for convenient importing into data analysis programs.These files are stored on an ASRC server and are available upon request.Meteorological data include air and dew point temperatures, wind speed, wind direction and barometric pressure.
Constancy of calibration procedures and activities are critical to the establishment of credible long-term data sets.As mentioned above, the summit ozone analyzer is calibrated and audited by NYS DEC using established procedures and following US EPA protocols.The original ozone analyzer was of the ethylene chemiluminescence type, and this method remains the Federal Reference Method for measurement of gaseous ozone.In 1985, the ozone analyzer changed to the ultraviolet absorption method, and has been this method since that time.The UV absorption method has been designated a Federal Equivalence Method for measuring gaseous ozone.Other gas analyzers are calibrated by ASRC staff using certified gas standards and dynamic gas dilution methods at least weekly.Certified standards include NIST traceable gas mixtures and/or certified permeation devices.Flowmeters used for calibrations are calibrated as needed, and at least annually.

TRENDS ANALYSIS
Annual box plots of O 3 , NO y , CO, SO 2 , and NO 2 mixing ratios at the summit observatory are shown in Figs.1-5, with data completeness represented by the box shading.The O 3 data makes up the longest data record, with reasonably complete data starting in 1976.A five year running mean trace is added to this box plot as a guide to the eye for this data set.There is some indication of year-to-year variability, but such excursions are small -generally 1-2 ppb in the mean value, and rarely more than 4 ppb.The earliest part of the data record is the exception to this relative slow variation, exhibiting a large increase (~10 ppb) in the late 1970's followed by equally large decrease through 1983, before slowly rising again in the late 1980's that lasted about 15 years.The relatively lower data completeness for those initial years may impact the mean values, although examination of the 1976 data showed the missing data was fairly well distributed throughout the year.The last ten years (2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014) have seen a decrease in the annual average ozone of about 4 ppb at the summit, as well as a decrease in the standard deviation of ozone values (Fig. S1).The decrease in standard deviation is due to the reduction of regional emitted ozone precursor gases (most notably NO x ), and the corresponding reduced frequency of summertime high concentration ozone events, as well as the increasing convergence of summertime and wintertime ozone levels which will discussed later.We will show later that the decrease in annual average is not spread equally across all seasons, but is most dramatic in the warm season.
This and numerous other long-term data sets for remotely observed tropospheric ozone are presented and discussed by Cooper et al. (2014).Whiteface is similar to the other eastern U.S. sites in that study for the period 1980-2010 with a slightly positive, but statistically insignificant trend in ozone.Like Whiteface, the eastern U.S. and European stations are more likely to exhibit downward ozone trends since 1990 or 2000, but many of these nascent trends are not statistically significant (Cooper et al., 2014).This can be contrasted with long-term sites in the western U.S. and mid-Pacific (namely Lassen National Park (CA), Joshua Tree National Park (CA), and Mauna Loa (HI)), where the ozone trend is increasing (and statistically significant) up to 2010.
The SO 2 trend shown in Fig. 2 provides a clear picture of the success, albeit sometimes in fits and starts, of Clean Air Act (CAA) regulation of major point source emissions of SO 2 (these sources are dominated by power plants or electricity generating units).After SO 2 concentrations at the summit observatory peaked in the early 1990's, there was a marked drop in the late 1990's after the phased in implementation starting in 1995 of the 1990 CAA Amendments Title IV Phase 1 reductions, followed by a flattening out over the next few years due to relatively high emission caps and trading of emissions allowances.Phase II of the CAA Amendments began in 2000, and another   S2 and more completely for the entire U.S. by Hand et al. (2012).On the whole, the measured ambient concentrations reflect the emissions from large electricity generating units in the Midwest and northeast states of the U.S. As shown in our companion paper (Schwab et al., 2016a), and in greater detail in Fig. S2 for the period 1990-2010 and S3 for the period 1994-2010, there is a strong association between the gaseous SO 2 measurements at the summit observatory, and the northeast U.S. emissions of SO 2 taken from Xing et al. (2013).These two quantities are correlated with an R 2 (coefficient of determination) of 0.88.
Fig. 3 shows that CO, also emitted from combustion sources, has also been trending down since the 1990's.Whereas SO 2 is mainly from electricity generating units, CO emissions are mostly from motor vehicles.This means that CO emissions cannot be as easily and completely regulated as SO 2 emissions, but the downward trend clearly indicates more efficient combustion in the areas upwind of Whiteface.We have not found a suitable explanation for the seemingly anomalous CO mixing ratios measured in 2000, but consider that year an outlier.
Oxides of nitrogen, NO y and NO 2 , shown in Figs. 4 and 5, have a quite different and complicated variation over time when compared to the previous three species.Fig. 4 shows that NO y mixing ratios at the summit observatory increased slowly through the 1990's and early 2000's, peaking in the mid-2000's.Large reductions in NO y concentrations were realized in the late 2000's, due at least in part to the US EPA's NO x Budget Program and the Clean Air Interstate Rule.These programs only account for a fraction of the total NO x emissions in the US, since mobile source account for nearly 60% of the NO x emissions in the latest 2011 National Emissions Inventory (http://www.epa.gov/ttnchie1/net/2011inventory.html).Although the increase that is showing up over the last three years for both NO 2 and NO y is difficult to fully explain, note that regulated NO x emissions have largely leveled out since 2009, and there was even a slight uptick in 2010 (http://www3.epa.gov/airmarkets/progress/reports/emissions_reductions_nox.html#figure1), which means that year to year variations could produce one or more years with higher concentration than the previous year.The three consecutive years of increased values would be unlikely even in this scenario, and no definitive explanation of this level of sustained increase has been found.Still, even with these recent puzzling results, the NO y trend over the last ten and even the last twenty years is downward, as would be expected from the emissions reductions in the U.S. required by EPA.
Gas concentrations measured at the Marble lodge site have a much shorter data record, beginning in earnest around 2000.An interquartile range box plot of the O 3 measurements at the Marble lodge site form 2001-2014 is presented as Fig. 6.Peak O 3 mixing ratios, as reflected in the 90 th percentile whiskers, show a statistically significant decrease over this period (p = 3.7 × 10 -5 ), while average O 3 mixing ratios have been relatively flat.

WIND SECTOR ANALYSES
An important tool for beginning to understand the source of pollutant gases to a location is the wind sector analysis technique, in which mean pollutant concentrations for a given set of wind direction ranges is displayed on a radial plot.What is important here is the notion of prevailing mean wind direction over relatively long transport distances (10's of km and more).Unfortunately, wind sensors located at the summit of Whiteface often experience strong local  perturbations and turbulence, due to mountain orography, radiative effects, and other flow disturbing phenomenon.For this reason we have elected to use 3-hour National Center for Environmental Prediction/North American Regional Reanalysis (NCEP/NARR) reanalysis data for the wind directions at Whiteface summit (Mesinger et al., 2006).Fig. 7 illustrates the pollution roses over eight wind sectors for the Whiteface summit trace gas measurements from the beginning of record through 2010.SO 2 has a highly skewed radial distribution of pollutants, NO, NO 2 and NO y distributions are moderately skewed, and CO and O 3 distributions are much closer to radially symmetric.Fig. 9 shows the change over time of the directional component of the SO 2 mixing ratio.For SO 2 the dominant contribution is consistently from the west and southwest sectors, but the change in magnitude is striking, and appears even more impressive than shown in Fig. 2. The largest SO 2 sources in the EPA power plant emissions database are to the southwest, as shown in Fig. S7 (http://www.epa.gov/airmarkets/progress/datatrends/index.html).The largest sources in this region have reduced their collective emissions by an average of greater than 90% between 1990 and 2014.Time evolution pollution rose plots for other gases are in the supplementary material (see Figs. S4-S6).

SEASONAL DEPENDENCE OF TRACE GAS CONCENTRATIONS
Monthly averaged O 3 mixing ratios for the ten-year period 2005-2014 measured at both the summit and lodge stations are presented in Fig. 9. Ozone is significantly higher in the spring as compared to all other seasons at these locations,   due in part to enhanced transport from the stratosphere (which is greatest in the spring, but only accounts for a very small increment at Whiteface -see Mohnen et al., 1977), and in part due to the enhanced spring production while NO x is still high and biogenic VOC's are entrained in a somewhat lower boundary layer than during summer.Ozone is lowest in late fall and early December as might be expected due to lower photochemical activity, but the higher mixing ratios in January and February do not fit the standard picture of higher ozone in the warm season.We speculate that very low boundary layer heights during winter, combined with snow enhanced albedo for photochemistry, may lead to higher than expected wintertime ozone production on a much smaller scale than that seen in the Unitah Basin in Utah (Ahmadov et al., 2014).Summit O 3 is systematically higher than lodge O 3 , from roughly two to as many as five ppb on average.A major contributor to this difference is the much greater surface deposition to the forested surroundings of the lodge site as contrasted with the rocky summit.Since NO x reacts with and titrates O 3 , the higher NO x at the Lodge could be a contributing factor for lower O 3 .Also the seasonal differences in the boundary layer as mentioned below likely impacts the O 3 .
Fig. 10 presents the seasonal variation of SO 2 , a primary pollutant gas, at both summit and Marble Lodge sites for the years 2005-2014.The Marble Lodge data exhibit a very strong seasonal dependence, with concentrations more than twice as high in the cold season months of January and February than in the warm season months of May, June, and July.This typically seen seasonal pattern is due to 1) more active photochemical removal in summer (both gas phase and aqueous), combined with 2) stronger combustion sources and 3) lower boundary layer heights in winter.The seasonal variation at the summit is much less pronounced with winter concentrations only about 50% higher than in summer.The more regional character of the mountaintop summit site helps to explain the smaller seasonal variation, since more long range transport, and therefore greater photochemical aging of the sampled air play much greater roles.The reaction rates are faster in the summer months (for both gaseous and aqueous oxidation) leading to lower concentrations in the warm season.However, wintertime local source and boundary layer height effects are much reduced, resulting in a weaker seasonal dependence.
In contrast to O 3 , these primary pollutants are systematically higher at the lodge site, which is more influenced by local sources and more "connected" to the surface boundary layer.The summit, in contrast, is much more likely to be separated from the surface layer, residing in the mixed layer and even at times the free troposphere.Hence the summit, on average, reflects more aged and oxidized air, with lower levels of precursor gases like SO 2 and NO x .
The seasonal dependence of O 3 measured at the summit observatory over the full data record is presented in Fig. 11, and as box plots in Figs.S8-S11.The most striking feature in this figure is the opposite behaviors of the summer and winter averages.The summer month average was just slightly higher than the spring average for the first part of the data record, until about 1990.Since then, the summer period has experienced steadily decreasing O 3 mixing ratios, while there is little change in the spring period.The difference in the mean O 3 for the spring and summer periods is currently greater than 5 ppb.Part or most of this summertime decrease can be attributed to the NO x SIP call program issued in 1998 and implemented in 2003 (Aleksic et al., 2013).Cooper et al. (2012) analyzed the O 3 data for this and many other rural North American sites for the period 1990-2010, and found similar results.At the same time, O 3 during the winter  months (DJF) has been slowly creeping up, approximately 2 or 3 ppb since the stable period in the late 1980's and early 1990's.From a simple p-value significance test, both the JJA and DJF linear trends are significant (p-values of 0.00005 and 0.006, respectively).However, the r-squared values of the trend lines are low (0.34 and 0.16), so these trends may not be considered "statistically meaningful" (Bryhn and Dimberg, 2011).The wintertime increase, if it continues and reaches a meaningful threshold, could indicate an increase in global or continental background ozone, due to global emission increases and/or changing climate (Lefohn et al., 1992;IPCC, 2013;Eyring et al., 2013;Shindell et al., 2013).

SUMMARY
Whiteface Mountain has been an active center of longterm research quality measurements of O 3 and related trace gases with year-round measurement activity beginning in the 1970's.The measurements of reactive trace gases occurs at two locations on the mountain, the summit observatory site and the Marble Mountain Lodge site; and is a collaborative effort between the SUNY Albany Atmospheric Sciences Research Center and the New York State Department of Environmental Conservation.This research activity has established long-term data sets on ozone and ozone precursors that provide an important window on the changing remote background atmospheric environment in eastern North America in the late 20 th and early 21 st centuries.The longest continuous data record is for ozone mixing ratio measured at the summit observatory, and this four decade time series shows an increase in the late 1970's, followed by a decrease and recovery in the 1980's.This was followed by a roughly 20 year period of little change, and indications of a decrease in starting around 2005.When examined as seasonal averages, the summertime summit O 3 concentrations exhibit a clear decrease, especially in the past decade.In contrast, wintertime O 3 has not decreased, and even shows a tendency toward increasing values.
Significant reductions in summit observatory SO 2 and CO concentrations (attributed to reduced emissions) have been observed over the past two and a half decades.NO y measured at the summit has decreased more modestly, and there is no clear trend (and even hints of a recent increase) in NO x concentrations.The same species measured at summit and lodge level shows similar seasonal variation of monthly averaged values.Ozone peaks in the spring, while primary pollutants (SO 2 , NO x , and CO) peak during winter.Ozone concentrations are systematically higher at the summit than the lodge, while the opposite is true for SO 2 and NO x .Analysis of pollution roses for the summit observatory indicates the largest pollution sources for this location are located in the west, southwest, and south sectors.

Fig. 1 .
Fig. 1.Whiteface Mountain Summit O 3 annual box plot for the period 1975 to 2014.The box encompasses the interquartile range of hourly values, and mean and medians are shown as horizontal line segments.

Fig. 2 .
Fig. 2. Whiteface Mountain Summit SO 2 annual box plot for the period 1992 to 2014.The details are as in Fig. 1.The years noted with the shaded bars labeled P1 and P2 are the implementation periods for Phase I and Phase II of the SO 2 emission reductions mandated by the Clean Air Act Amendments of 1990.

Fig. 3 .
Fig. 3. Whiteface Mountain Summit CO annual box plot for the period 1989 to 2014.The details are as in Fig. 1.

Fig. 4 .
Fig. 4. Whiteface Mountain Summit NO Y annual box plot for the period 1989 to 2014.The details are as in Fig. 1.

Fig. 5 .
Fig. 5. Whiteface Mountain Summit NO 2 annual box plot for the period 1994 to 2014.The details are as in Fig. 1.

Fig. 6 .
Fig. 6.Marble Lodge O 3 annual box plot for the period 2001 to 2014

Fig. 7 .
Fig. 7. Whiteface Mountain Summit mean directional trace gas mixing ratio for each trace gas from the beginning of record to 2010.Wind direction retrieved from 3-hour NCEP reanalysis data.

Fig. 8 .
Fig. 8. Whiteface Mountain Summit mean directional SO 2 mixing ratio for three time periods from 1989 to 2010.

Fig. 9 .
Fig. 9. Whiteface Mountain Summit and Lodge O 3 averaged by month for the consolidated period 2005 to 2014.

Fig. 10 .
Fig. 10.Whiteface Mountain Summit and Lodge SO 2 averaged by month for the consolidated period 2005 to 2014.

Fig. 11 .
Fig. 11.Whiteface Mountain Summit annual and seasonal mean O 3 time-series smoothed using a 5-year running average.

Table 1 (
a). List of gaseous parameters, methods, and instruments for current measurements at the Whiteface summit observatory.

Table 1 (
b). List of gaseous parameters, methods, and instruments for current measurements at the Marble lodge site.

Table 1 :
ASRC -Atmospheric Sciences Research Center; DEC -Department of Environmental Conservation (New York State).