Estimates of Lightning NOx Production Based on OMI NO2 Observations over the Gulf of Mexico

In this study we evaluate nitrogen oxide (NOX= NO + NO2) production from lightning in convective storms over the Gulf of Mexico region first during August 2008 and July 2011 and then for all summer months during the 2007 to 2011 time period. We use data from the Ozone Monitoring Instrument (OMI) aboard NASA's Aura satellite along with lightning data from the World Wide Lightning Location Network (WWLLN), a ground based lightning detection system. A special algorithm was developed to retrieve the lightning NOx (LNOx) signal from OMI. The algorithm in its general form takes the total slant column NO2 from OMI and removes the stratospheric contribution and tropospheric background (NO2 from sources other than lightning) and includes an air mass factor appropriate for the profile of lightning NOX to convert the slant column LNO2 to a vertical column of LNOx. The WWLLN flash counts are adjusted based on WWLLN detection efficiencies derived through comparisons between WWLLN and satellite-based lightning observations. The analysis is conducted on a 1o x 1o grid covering the Gulf of Mexico and immediately surrounding land areas. WWLLN flashes are totaled over a time period (based on climatological winds) in which an air parcel is expected to remain in a grid box (3-hours for the Gulf of Mexico region) prior to OMI overpass. Grid cells are restricted to those containing flash counts greater than particular threshold values and those having pixels with cloud radiance fraction (CRF) greater than a criterion value (0.9) indicative of deep, highly reflective clouds. Therefore, the method discussed in this presentation is appropriate over regions of active or recently active convection. With this criterion, the tropospheric background is minimal. We use summation and regression methods to estimate the monthly mean LNOx production per flash over the region. For the summation method we conclude that the best estimates of LNOX production in active convection in the Gulf of Mexico come from analyses using flash thresholds of 300 and 500 flashes. Averaging over the two months considered (August 2008 and July 2011) and over the results for these two flash thresholds, we obtain a best estimate of 233 moles per flash over this region using the summation method and 167 moles per flash using the regression method. Expansion of the analysis to all summer months during 2007 – 2011 yields a monthly mean LNOx production of 267 ± 73 moles per flash for the summation method and 196±63 moles per flash for the regression method. These results are consistent with literature estimates and more robust than many prior estimates of LNOx production due to the large number of storms considered.