Gail Tonnesen, EPA Region 8 AQRS Meeting, May 18, 2017 Update on Progress in Monitoring and Modeling Studies to Characterize Sources of Ozone and Regional Haze for Air Quality Planning Gail Tonnesen, EPA Region 8 AQRS Meeting, May 18, 2017

Air Quality Planning Context Second round of Regional Haze SIPs are due in 2021, third round due in 2028 and every subsequent 10 years: Regional Haze Rule sets a goal of achieving natural visibility conditions at Class I areas by 2064. States must demonstrate reasonable progress toward the natural visibility goal. Need more research to estimate natural visibility and the contributions of international transport to regional haze. Ozone SIPs for the 70 ppb NAAQS due in 2020 or 2021. Areas that fail to attain will be bumped up and will require additional modeling and new SIPs after 2020. Planning efforts continue for areas that did not attain the 2008 75 ppb O3 NAAQS and that are being bumped to more serious non-attainment status. Evaluation of interstate transport contributions to O3 are needed to support states “good neighbor” transport SIPs. PM2.5 SIP planning efforts are ongoing. Exceptional Event demonstrations for ozone and PM are ongoing. Near term research will benefit O3 and Haze SIPs due in 2020 and 2021, longer term research will benefit future planning efforts.

2015 AQRS Meeting: Strategies for Improving the State of the Science for modeling O3 in the western US More monitoring data to improve characterization background O3 and to evaluate the accuracy of model-based estimates of USB: More measurements to improve characterization of vertical O3 profiles. Network of O3 LIDAR vertical profiles (NASA TOLNET pilot study) More ground based O3 and precursor measurements in rural areas. Perform comprehensive model evaluation studies using new monitoring data to assess contributions to background O3. Do global models accurately estimate BC inflow? Do regional models accurately simulate natural sources of O3 from wildfires and biogenic precursors? Do regional models accurately simulate vertical mixing of O3? Need projections of future trends in global O3. Increase state/federal & planner/researcher collaborations to improve modeling and data analysis for O3 transport, wildfires, and stratospheric intrusion.

Summary of recent and planned work Progress in Stratospheric Intrusion Exceptional Events Special Field Studies: LVOS, FAST-LVOS, CABOTS Modeling and analysis of FRAPPE/DISCOVER-AQ Field Study Analysis of trends in Ozone Regional Haze Modeling and Visibility conference Planned Future Work: FIREX, FIRECHEM studies Studies of background O3 HAQAST projects TOLNet progress and future work TEMPO Satellite launch

Stratospheric Intrusion Workgroup Workgroup meeting semi-monthly since 2012, include federal and state research and air quality management communities. Lin et al. (2015) show “more frequent late spring stratospheric intrusions when the polar jet meanders towards the western United States, such as occurs following strong La Niña winters.” Few stratospheric O3 events in 2015 and 2016 (during El Niño conditions) but winter 2016/17 had La Niña conditions, and 2017 has already had several possible intrusion events with maximum daily 8-hour (MD8) averages approaching or exceeding the NAAQS: March 18: 95 ppb (87 ppb regulatory) at Mt Washington NH. April 9: 74 ppb in Colorado Springs. April 22: 82 ppb Gothic, CO; 80 ppb Navajo Lake, NM; 75 ppb Centennial, WY; and five others sites in UT, CO and NM exceed 70 ppb. April 23: 71 ppb NREL, CO; 71 ppb Rocky Flats, CO Several additional days with MD8 between 65 to 69 ppb.

Example plots for SI evaluation GFS 04/22/17, 18Z initialization of total column ozone, and 600 mb Relative Humidity (RH) and Isentropic Potential Vorticity (IPV) Twice daily atmospheric soundings http://weather.uwyo.edu/upperair/sounding.html Black: IPV >1, Pink: RH < 10%

Example plots for SI evaluation RAQMS 04/22/17 12Z simulation, 6-hour forecast of ozone at 3km RAQMS 04/22/17 12Z run, 12-hour forecast of ozone at 40N cross-sectional

Example plots for SI evaluation Rapid Refresh Model (RAP) Simulations AIRS Satellite Derived Total Column CO

Resources for SI O3 Analyses NOAA RAQMS simulations and IDEA Trajectory analysis tool (Brad Pierce, NOAA). Tropospheric Ozone Lidar Network (TOLNet) ground based lidars (Mike Newchurch, UAB; Andy Langford et al., NOAA/ESRL) Atmospheric Sounding, website hosted by Univ of Wyoming AirNow surface O3 monitoring by states, tribes and EPA Additional surface O3 sites by NOAA/GMD, NPS, FS and others. NOAA Rapid Refresh (RAP) model simulations: https://rapidrefresh.noaa.gov/ Satellite based measurements of ozone, CO, and aerosol optical depth. NASA AJAX aircraft observations. Staff at state environmental agencies with expertise in meteorological analysis and local conditions. SI Workgroup is an example of federal and state staff working together to synthesize data and resources from multiple agencies and programs to address an important air quality planning need.

Background O3 Scientific Assessment Western States Air Resources Council (WESTAR), Western Regional Air Partnership (WRAP), and American Petroleum Institute (API) are co-sponsoring a scientific assessment on background O3 and future research needs. The assessment is lead by Prof. Dan Jaffe, includes academics, NOAA and EPA staff, and will consider current research to examine: 1. Sources of background O3 2. Background O3 as seen by observations 3. Background O3 as seen by models 4. Reconciling observations and models 5. Temporal and spatial variations in background O3 6. Research needs to improve our understanding of background O3. The assessment is focused on technical and scientific aspects of non-controllable O3 sources (NCOS) that are relevant to policy, but will not directly address policy. Broad participation by the O3 research community and stakeholders is encouraged. The end result will be one or more peer-reviewed journal articles incorporating the results and discussions from the workshop.

Key questions raised at March 28-29 Denver Workshop Trends in O3 are complex and depend on location and period of record. Can we reconcile observed trends with changes in domestic and international emissions, climate, wildfires, etc.? Explain why are there strong downward trends in California, but much weaker trends at other western sites. Regional/urban scale modeling of O3 depends critically on boundary conditions, so propagation of boundary uncertainties can cause significant problems. How can we improve boundary conditions for regional modeling? USBO includes stratospheric influence, biogenic precursors, wildfires and international transport. How can we improve our estimates of USBO on a daily basis? Concern that models do not accurately predict wildfire contributions to O3. All models have uncertainties and biases. Can we develop an estimate of uncertainty for model derived estimates of USBO? What are the assumptions made in using the “relative response” method to correct for model bias, and are these assumptions valid for addressing model bias in USBO? Current satellites have little sensitivity to key species in the boundary layer. How can we better utilize current and future satellite data to improve estimates of O3 sources? What is the sensitivity of USBO to international emission changes? How well do we understand local/controllable contributions to O3?

BG O3 Assessment: broad, draft recommendations Continue to improve observation network: Low cost improvements Improvements requiring larger investments Better coordination within and between experimentalists and modeling groups Improved use of existing data to evaluate our understanding of USBO, e.g. statistical relationships between tracers and meteorological conditions, etc. Targeted studies on background O3 Continue to improve models Need better evaluation of models with observations that are most relevant to the application; Need for planning and integration of future satellite observations Further development of integrated air quality management systems.

Special Field Studies for O3 FRAPPE/DISCOVER-AQ 2014 Field Study in Colorado – modeling and analysis are in progress. NCAR, NASA, Colorado Dept of Public Health and Environment. Las Vegas O3 Study 2013, NOAA, and Clark County NV Dept of Air Quality, several publications describe contributions of stratospheric O3, wildfires and transport to O3 in southern NV. Fires, Asian and Stratospheric Transport Las Vegas O3 Study (FAST-LVOS) in 2017 will address contributions to O3 from southern California and from wildfires, transport of O3 aloft and how it reaches the surface in the Las Vegas Valley. California Baseline Ozone Transport Study (CABOTS) – 2016 Field Study address important scientific questions on O3 transport and air quality in California, includes NOAA, NASA, Universities, and California Air Resources Board: https://www.arb.ca.gov/research/cabots/cabots.htm NASA Health and Air Quality Sciences Team (HAQST) studies beginning 2017: Supporting the use of satellite data in State Implementation Plans (SIPs), led by Arlene Fiore (Lamont-Doherty Earth Observatory/Columbia University) High resolution particulate matter data for improved satellite-based assessments of community health, led by Pat Kinney (Boston University) Improved NEI NOx emissions using OMI Tropospheric NO2 retrievals, led by Brad Pierce More resources are needed to support collaborations among state, federal and academic researchers for modeling and analysis of field study data.

Recent Assessment of O3 Trends Tropospheric Ozone Assessment Report (TOAR): Led by Owen Cooper at NOAA, eight peer reviewed articles on topics related to background O3: http://www.igacproject.org/activities/TOAR Critical review of the present-day and near-future tropospheric ozone budget Tropospheric ozone observations Global ozone metrics for climate change, human health, and crop/ecosystem research Present-day ozone distribution and trends relevant to human health Present-day ozone distribution and trends relevant to vegetation Present-day ozone distribution and trends relevant to climate and global model evaluation Assessment of global-scale model performance for global and regional ozone distributions, variability, and trends Database and metrics data of global surface ozone observations Key findings and recommendations for future research

Wildfire Research Studies EPA published guidance in 2016 on Wildfire O3 Exceptional events. Integrated planning effort for three field studies in 2019: Fire Influence on Regional and Global Environments Experiment (FIREX, NOAA) FIREChem (NASA) Fire and Smoke Model Evaluation Experiment (FASMEE, Joint Fire Science Program) Colorado State University wildfire study is proceeding in 2018. These studies will be valuable for assessing wildfire contributions to O3 regional background, O3 exceptional events, and regional haze. Can we also leverage these studies to learn more about background O3 and haze conditions in the absence of fire? Majority of monitoring will be in smoke plumes, but measurements upwind of plumes may be valuable for characterizing conditions in the absence of fire.

Regional Haze Modeling Clean Air Act goal is to achieve natural visibility conditions at Class I areas by 2064 with linear progress in reducing haze (in deciviews) on the worst 20% days and no degradation on the best 20% days. Uniform rate of progress (aka Glidepath) is defined as the slope of the line from baseline worst 20% deciviews to the natural deciviews. Analysis of IMPROVE monitoring data is used to estimate natural visibility and to track actual progress in reducing haze. Visibility trend in IMPROVE data at Sawtooth Wilderness Area, ID Model simulations are used to identify sources that contribute to haze and to estimate progress in reducing in reducing haze for each 10 year planning period. If model estimates of progress are above the glidepath, states must demonstrate that no other controls are reasonable. States are not responsible for visibility impairment caused by international transport, but estimates of international transport have large uncertainty which makes it difficult to estimate the contributions of US sources versus international sources to visibility impairment.

Uncertainty in Global CTMs for Haze Planning Previous modeling studies indicate that international transport and natural background are large contributors to ozone and regional haze at some western sites. The Clean Air Act and Regional Haze Rule provide relief for cases in which ozone and PM2.5 exceedances and visibility impairment are caused by international emissions, but model estimates for these contributions are uncertain: Global models have bias and error in seasonal performance. Global scale models have not been well evaluated for specific episodes relevant to regulatory planning. International anthropogenic emissions are treated differently than natural emissions in some regulatory applications, so source attribution results are needed to distinguish international transport caused by anthropogenic emissions versus natural sources.

Lassen Volcanic National Park: CAMx source contributions to sulfate For regional haze planning, sulfate concentrations less than 1 ug/m3 are important (1 ug/m3 causes light extinction of about 8 Mm-1 or 50 miles reduction in visibility compared to natural conditions). How much confidence do we have in global model estimates of boundary conditions? What is the contribution to sulfate from natural sources and international anthropogenic SO2 emissions?

CAMx performance for sulfate and nitrate: Lassen Volcanic National Park 2011, daily 24-hour average Biased high in winter Biased high for wildfire Biased low in spring Biased high for wildfire Biased high in winter and spring

Regional Haze Research List of key research needs identified at the 2016 AWMA Visibility Conference: Need for updated estimates of natural haze conditions. Improved data analysis methods to exclude days influenced by wildfires. Estimates of international transport for natural vs. anthropogenic PM. Evaluation of global models. Several Regional Haze Modeling Studies planned or in progress: EPA Regional Haze Source Apportionment modeling for 2028. EPRI Regional Haze modeling study will assess the contributions of international transport to regional haze at U.S. Class I areas in 2028, including simulations using the latest versions of the GEOS-Chem global chemistry model and the CAMx regional photochemical grid mode. WRAP work plan to develop an updated base year visibility modeling platform and projection for 2028.

Limitations of Global Models Global models are not expected to perform well in urban areas because of coarse spatial resolution (0.5 to 2 degrees), so global model evaluation is limited to rural and remote areas that do not have strong spatial gradients in emissions. Air quality planning relies on high resolution regional models that are nested within a global model. Because of uncertainty in international emissions and coarse spatial resolution, there may be limits on the accuracy of global models. Need to consider methods to identify and accommodate uncertainty and errors in the global models, e.g., exclude days from planning analysis for which the coupled global/regional modeling system has poor performance. Even with errors and bias, current global models are a major improvement compared to past practice of static boundary conditions.

Key Research Needs for Global/Regional Models More detailed evaluation for individual monitors and days: Evaluation using monthly means or averages for all monitors are not adequate. More ground based O3 and precursor measurements in rural areas. Perform comprehensive model evaluation studies using new monitoring data to assess contributions to background O3 and PM2.5 species. Do global models accurately estimate BC inflow? Do regional models accurately simulate natural sources of O3 from wildfires and biogenic precursors? Do regional models accurately simulate vertical mixing of O3? Need projections of future trends in global O3. Perform model source attribution studies to evaluate natural and international contributions to O3, PM2.5 and regional haze. Evaluate effects of model vertical resolution and methods used to represent vertical advection and diffusion. Increase state/federal & planner/researcher collaborations to improve modeling and data analysis for O3 transport, wildfires, and stratospheric intrusion.

Potential opportunities for collaboration Western Air Quality Workshop in Sept 6-8, 2017 in Boulder Co will review state of the science and recommend formation of inter-agency workgroups: meteorology in complex terrain. Evaluation of boundary conditions derived from global CTMs. Improved estimates of oil and gas emissions. NASA Health and Air Quality Applied Science Team (HAQAST) Partner atmospheric scientists with AQ managers to leverage earth science tools Continuing modeling and analysis of DISCOVER-AQ and FRAPPE studies. Expansion of TOLNet O3 lidar effort to evaluate contributions to O3. FIREX/FIREchem/FASMEE wildfire studies. Planning for use of TEMPO satellite data in 2020. Others?