1. Importance of Lightning NO for Regional Air Quality Modeling Thomas E. Pierce/NOAA Atmospheric Modeling Division National Exposure Research Laboratory U.S. Environmental Protection Agency Research Triangle Park, North Carolina (pierce.tom@epa.gov) Presentation to the Joint Action Group for Lightning Detection Systems Office of the Federal Coordinator for Meteorology Silver Spring, Maryland December 3, 2003

2. Background Regional air quality models are gridded 3D realizations of physical and chemical processes. They are used for simulating photochemical smog, acid deposition, visibility, fine particulates, mercury, and other air toxics. In addition to chemical and meteorological modules, they must consider hourly estimates of anthropogenic and natural sources of emissions. The global climate modeling community widely recognizes the importance of lightning for generating nitric oxide (NO). The IGAC/Global Emissions Inventory Activity (GEIA) distributes a global inventory of lightning NO at 1-deg resolution. With a few exceptions (noted on the next slide), the regional air quality modeling community has given scant consideration to lightning NO, primarily because of the difficulty in obtaining lightning strike data and the uncertainty with the NO emission factors.

3. Background (continued) Selected references related to regional air quality and lightning NO: Logan, J. (1983) Nitrogen oxides in the troposphere: global and regional budgets, JGR, vol. 88, pp. 10785-10807. Pickering, K. et al. (1990) Model calculations of tropospheric ozone production potential following observed convective events, JGR, vol. 95, pp. 14049-14062. Novak, J. and T. Pierce (1993) Natural emissions of oxidant precursors, Water, Air & Soil Pollution, vol. 67, pp. 57-77. Biazar, A. and R. McNider (1995) Regional estimates of lightning production of nitrogen oxides, JGR, vol. 100, pp. 22861-22874.

4. Introduction Estimates of NO produced by lightning have been included in the Regional Acid Deposition Model (RADM). In a preliminary study, the influence of this NO source on regional photochemistry has been investigated, and differences in resultant O 3 concentrations have been examined. Lightning NO emissions vary temporally & spatially (horizontal & vertical), as well as in source strength (emission intensity). U.S. National Lightning Detection Network (NLDN) data and literature values have been used to estimate these emissions for the eastern USA. RADM simulations for 25 continuous days have been performed. WHATImpact of lightning nitric oxide (NO) emissions on regional photochemistry WHERE Eastern USA WHEN July 19 - August 13, 1988

5. Lightning NO Emissions Source: National Lightning Detection Network (NLDN) July 19-August 13, 1988. Only cloud-to-ground (CG) flashes. Detection efficiency assumed to be 70%. CG flashes: Flash assigned to corresponding grid cell. Each CG flash produces 33 kg NO (Price et al., 1996). NO distributed from ~5000 m (layer 17) to the surface. IC Flashes: Assumed 2.7 IC flashes per CG flash (Price and Rind, 1993). Each IC flash assumed to produce 1/10 as much NO as a CG flash. NO distributed in likely cloud depth (model layers 17-19).

14. Time Step Layer 1

15. Layer 10

16. Layer 17

17. Ramifications for Local O 3 For this RADM simulation*, lightning NO caused average surface O 3 to rise by 2 ppb. Differences were very localized... On 250 occasions, lightning NO caused O 3 to rise above 120 ppb. On 240 occasions, O 3 concentrations >80 ppb increased by >10 ppb. On 1550 occasions, O 3 concentrations >60 ppb increased by >10 ppb. *Simulation is 216h x 33 horz grids x 38 vert grids x layer 1 = 256,608 “occasions.”

18. Conclusions Across the eastern U.S. during the summer, lightning provides a highly variable source of NO, which can be defined in space and time with the National Lightning Detection Network (NLDN). In a photochemical model simulation of July 19-August 13, 1988, lightning NO accounted for 11% of total NO x emissions, raised average surface ozone by 2 ppb, and produced a maximum increase of 33 ppb. Recommended model refinements include more accurate NO production rates and improved handling of the vertical distribution and transport of lightning NO.

19. Recommendations The impact of lightning NO on regional photochemistry needs to be reassessed with an updated atmospheric chemical transport model, such as the Community Multiscale Air Quality (CMAQ) modeling system. A reliable and inexpensive source of lightning data for regional air quality modeling needs to be identified, or a reliable parameterization of lightning flashes needs to be developed.