Long-Term Trends in Reactive Nitrogen Deposition in the United States

I. Introduction

Long-term monitoring of ambient air quality and deposition is necessary to characterize trends in human and ecosystem exposure and to gauge the effectiveness of air pollution control programs. Such datasets are rare because of the difficulty and capital required to consistently and accurately collect and analyze samples over time from a spatially adequate number of regionally representative sites. Most of the national air pollutant monitoring networks producing these datasets were established in the 1970s and 1980s and focused on the human health-based National Ambient Air Quality Standards (NAAQS) criteria pollutants (e.g. sulfur dioxide (SO₂), nitrogen dioxide (NO₂), ozone (O₃), particulate matter < 2.5 μm (PM_2.5)) or reporting acid rain trends and visibility impairment.

Under Title IV of the Clean Air Act Amendments (CAAA), electric generating units (EGU) were required to make significant reductions in emissions of SO₂ and oxides of nitrogen (NO_x; i.e. nitric oxide (NO) and nitrogen dioxide (NO₂)). While NO_x has continued to be regulated under stationary and mobile emissions programs (e.g. NO_x Budget Trading Program), reduced nitrogen (NH_x; i.e. particulate ammonium (pNH₄) plus gaseous ammonia (NH₃)) remains unregulated despite its contributions to PM_2.5 formation and total reactive nitrogen (Nr) deposition. Several long-term monitoring networks have measured components of Nr species for several decades (Table 1).

Table 1.

Existing monitoring networks in the US that measure components of reactive nitrogen in the atmosphere or precipitation (wet deposition). Data from these networks are used by state, local and federal agencies, researchers, and industries to assess trends in atmospheric pollution and deposition.

Network	Reactive Nitrogen Measurements	Measurement Interval	Website
Clean Air Status and Trends Network (CASTNET)	Ambient concentrations of pNH4+, particulate nitrate (pNO3−), nitric acid (HNO3)	weekly	https://epa.gov/castnet
National Atmospheric Deposition Program (NADP) National Trends Network (NTN)	Concentrations of NO3− and NH4+ in precipitation; precipitation amounts	Weekly	http://nadp.slh.wisc.edu/NTN/
NADP’s Atmospheric Integrated Research Monitoring Network (AIRMoN)	Concentrations of NO3− and NH4+ in precipitation; precipitation amounts	Daily (event based)	http://nadp.slh.wisc.edu/AIRMoN/
NADP’s Ammonia Monitoring Network (AMoN)	Ambient concentrations of NH3	Bi-weekly	http://nadp.slh.wisc.edu/AMoN/
Chemical Speciation Network (CSN)	Ambient concentrations of pNO3−, pNH4+	Daily (1:3 or 1:6 day)	https://www3.epa.gov/ttn/amtic/speciepg.html
Interagency Monitoring of Protected Visual Environments (IMPROVE)	Ambient concentrations of pNO3−, particulate nitrite (pNO2−)	Daily (1:3 day)	http://vista.cira.coloradostate.edu/improve
National Core (NCore) Multipollutant	Concentrations of NO, total oxidized nitrogen (NOy); PM	Hourly	https://www3.epa.gov/ttn/amtic/ncore;
Network; State and Local Air Monitoring Stations (SLAMS); National Air Monitoring Stations (NAMS)	speciation (CSN or IMPROVE)		https://www3.epa.goV/airquality/montring.html
Photochemical Assessment Monitoring Stations (PAMS)	Concentrations of NO, NOy, NOx	Hourly	https://www3.epa.gov/ttn/amtic/pamsmain.html
Near-road NO2 Monitoring	Concentrations of NO2	Hourly	https://www3.epa.gov/ttnamti1/nearroad.html

While monitoring data are used to assess regional long-term trends in air concentrations¹ and wet deposition of some Nr species², the NADP Total Deposition (TDep) measurement-model fusion method³ is widely used for assessing trends in total (wet + dry) Nr deposition in the U.S. Briefly, the TDep method combines measured concentrations and wet deposition with modeled values where measurements are lacking (spatial gaps or unmeasured species).

Here we use measurements, TDep products, and emission inventories to discuss current trends in atmospheric concentrations and deposition of Nr and their relationship to trends in emissions. This analysis identifies several examples where improvements in monitoring, modeling, and emissions inventories are needed to better characterize the linkages between trends in emissions and changes in the atmospheric composition of Nr.

II. Assessment of current monitoring trends in emissions, ambient concentrations, and deposition

Annual emissions of SO₂ and NO_x have decreased substantially (−83% and −57%, respectively) from the period from 1990–1992 to 2014–2016 (Figure 1). This is attributable to EGU controls (EGU emission reductions of −85% for SO₂ and −77% in NO_x)⁴, market-driven changes to cleaner fuels, and mobile source controls (mobile source reductions of −89% for SO₂ and −46% in NO_x)⁴ (Figure 1).

Figure 1.

Three-year averaged annual emissions trends for SO₂, NO_x, and NH₃ by source category.⁴ Source categories are grouped as regulated EGUs, transportation, and other which includes everything from fuel combustion from non-EGU sources, industrial processes, agricultural sources and events such as wildfires and prescribed burns. The percent decrease in overall emissions is noted above the 2014–16 bars.

This decline is reflected in the long-term monitoring of ambient concentrations over the same period. The decreasing SO₂ concentration trend measured at eastern CASTNET sites (−86%; Figure 2; summarized in Table 2) showed a linear relationship between EGU emissions and ambient concentrations (R²= 99%).⁵ Data that support linkages between emissions and environmental results (i.e. SO₂) provide accountability for regulators and the regulated community.

Figure 2.

Trends in annual aggregate mean SO₂ (top), total nitrate (middle), and pNH₄ (bottom) concentrations from CASTNET eastern reference sites.⁷ The CASTNET reference sites are split into eastern and western regions due to the spatial density of the measurement sites, concentration differences, a difference in filterpack collection flow rate, and different start dates of operation. Only eastern sites are discussed as they are in closer proximity to EGUs and more reflective of the trends.

Table 2.

Summary of percent differences for oxidized sulfur, oxidized nitrogen, and reduced nitrogen in emissions, concentrations, and total deposition over different time periods of comparison. All percent differences are obtained from three-year averages at the beginning and end of the time period as indicated.

Species	Time Period	Emissions	Concentrations	Total Deposition
SO2	1990–92 to 2014–16	−83	−86	-
	2000–02 to 2014–16	−76	−80	−58◊
NOy	1990–92 to 2014–16	−53ζ	−48¥	-
	2000–02 to 2014–16	−48ζ	−48¥	−35
NHx	1990–92 to 2014–16	−19†	−63‡	-
	2000–02 to 2014–16	−15†	−58‡	30
	2008–10 to 2014–16	−17†	−39‡; 24†; 7 ± 2†δ	19

^◊-reported as total S;

^ζ-reported as NO_x;

^¥-reported as total nitrate;

^†-reported as NH₃;

^‡-reported as pNH₄⁺;

^δ-Butler et al., 2016

Deriving this type of relationship between emissions and concentrations is more convoluted for Nr species. Atmospheric processing converts the NO_x emitted by sources (reported by emissions monitors) to a diverse number of oxidized N compounds (NO_y) which monitoring networks either measure as total NO_y (by chemical conversion of all NO_y compounds to NO prior to detection) or as a fraction of NO_y (e.g. filter-based methods report out on total nitrate [the sum of nitric acid (HNO₃) and particulate nitrate (pNO₃]). Also, NO_x emissions are more distributed across source types (e.g. 26% EGUs, 52% transportation, 22% other in 1990)⁴ (Figure 1). Large decreases in NO_x emissions during the period from 1990–1992 to 2014–2016 are reflected in a marked decrease in ambient concentrations of total nitrate (−48%) at CASTNET eastern reference sites (Figure 2), and in NO₂ satellite observations.⁶

For NH_x species, emission sources emit NH₃, which in the atmosphere can readily convert to pNH₄ or remain as NH₃ depending on meteorological conditions and availability of acidic pollutants as precursors to pNH₄. Monitoring networks need to measure both forms to accurately represent NH_x. CASTNET ambient pNH₄ concentrations show a similar decreasing trend (−63%) as those reported for SO₂ and total nitrate (Figure 2), yet NH₃ emissions have decreased at a much slower rate (−19%) since 1990–1992.⁴ Measured annual ambient NH₃ concentrations at 21 NADP/AMoN sites with long-term sampling records increased 24% from 2008–2010 to 2014–2016 (Figure 3). An increasing NH₃ trend (7 ± 2%) was also identified in a study on similar sites that accounted for variability in seasonality and regional location.⁸

Figure 3.

Trend in annual aggregate mean NH₃ concentrations from 21 NADP/AMoN sites.

Trends in total (wet + dry) deposition in the U.S. were derived from TDep results and should be reflective of those for emissions and ambient concentrations. Total Sulfur (S) deposition decreased −58% from 2000–2002 to 2014–2016 and total NO_y deposition decreased −35% over the same period, showing significant but less dramatic trends than measured concentrations (Figure 4; summarized in Table 2). However, total NH_x, deposition increases by 30% over the same time period, and comprises a decrease in dry pNH₄ deposition (−17%) and increases in wet NH₄⁺ deposition (+24%) and dry NH₃ deposition (+54%), which contribute 2%, 30% and 18% to the total Nr budget, respectively (Figure 4). Ambient NH₃ can be entrained in precipitation, thus higher NH₃ concentrations likely explain the observed increases in wet NH₄⁺ deposition. The decreasing pNH₄ concentrations and the increasing NH₃ concentrations suggest that less of the NH₃ emissions are partitioning to the particle phase which is supported by the concurrent decline in SO₂ and NO_y emissions and concentrations which reduces the potential for acidic pollutants to react with gaseous NH₃ and convert to PM.⁸ Summing the concentration averages for pNH₄ and NH₃ over the period from 2008–2010 to 2014–2016 and calculating the difference provides a rough estimate of NH_x concentration trend (−4%) which is more proximate to the NEI NH₃ emissions trend of −17% over this period, though still a substantial difference.

Figure 4.

Trends in Nr deposition output by the TDep measurement-model fusion method. Top plot is the deposition flux of total Nr and its oxidized and reduced components (kg-N ha⁻¹). The lower plot is the percentage of total Nr deposition for each modeled species and its deposition pathway.

The TDep NO_y total deposition maps from 2000–2002 to 2014–2016 show that the reductions in NO_y deposition have been significant downwind of large EGU sources in the Eastern U.S. (Figure 5). Urban areas are now easily identified as the major NO_y hotspots. The total NH_x deposition map shows increases in agricultural source regions (e.g. midwest U.S., eastern NC, southeastern PA) (Figure 5). The total Nr deposition predicted by the TDep method is now approximately half NO_y and half NH_x (Figure 4). This trend has also been observed in other studies.^2,10

Figure 5.

TDep method deposition maps of NO_y and NH_x from 2000–2002 to 2014–2016.⁹

III. Limitations to trends analysis and needed research

The trends analyses in Nr emissions, concentrations, and deposition described in the previous section are not without limitations. Linking trends in emissions and atmospheric concentrations for Nr species is not as straight-forward as S, as there are more Nr species, more reactivity, and more sources to convolute these linkages. Emissions inventories for non-EGU sources are not robust and improvements are needed for all Nr species. These sources have greater uncertainties as they are more variable with time (e.g., agricultural and biogenic sources), are episodic (e.g. wildfires), and are typically calculated via mass-balance techniques.¹¹ Studies suggest current inventories for mobile emissions are overestimated for NO_x^12,13,14 and underestimated for NH₃.^15,16 Agricultural sources (e.g. livestock production, emissions from fertilized soils) account for 80% of U.S. NH₃ emissions^4,17,18 and are poorly characterized by agricultural practice and activity data in emissions inventory development.¹⁹

There are substantial limitations to the available Nr concentration measurements and how those measurements are used to assess total deposition. The TDep methodology does not utilize measured NH₃ concentrations because of a non-linear relationship with the modeled bi-directional deposition velocities. Also, existing network measurements for NO₂ (e.g. U.S. AQS), are also not currently utilized. Planned newer versions of the TDep method will address these limitations in the near future. Approximately 13% of the total Nr deposition budget is either not measured or not utilized by the TDep method (Figure 4). A fraction of this is organic nitrogen (ON) which is uncharacterized.

Research needed to address limitations

Routine Nr monitoring could be expanded to include bulk sampling of ON in precipitation and PM to develop more complete Nr budgets. Additionally, low-cost passive samplers for NH₃ and NO₂ could be added to existing networks to help characterize gradients from urban and agricultural areas to rural, non-source impacted areas. This could be conducted in tandem with satellite assessments to identify new monitoring locations and to better understand measurement spatial representativeness. Further development of low-cost methods for directly measuring dry deposition, suitable for routine network operation, is also a high priority. Finally, there are constant improvements in the accuracy of chemical transport models (CTMs) used to develop long-term time series of concentrations and deposition. These new estimates need to be reconciled with older estimates, especially for trends assessment where consistency is essential.

Satellite measurements of tropospheric NO₂ and NH₃ concentrations can augment current monitoring and modeling strategies for Nr and address some of these limitations. Satellite data products have been used to quantify regional and point-source scale emissions,^20,21 including episodic emissions (e.g. wildfires)^22,23 to improve emissions inventories.^24,25 Also, satellite-derived long-term trends for concentrations of NO₂^26,27 and NH₃^28,29,30 support surface monitoring trend data and provide information on spatial variability³¹ not achievable with surface networks. Satellite data products have been used in conjunction with measurements, CTMs, and deposition models to estimate trends in Nr deposition^32,33 or to evaluate and improve the CTMs^28,34,35,36 and thus improving modeled deposition estimates providing more accurate trends.