Use of Tropical Ground-Based Lightning Data for Improving Global Parameterized Lightning Fields in Chemical Transport Models

Lightning is an important natural source of reactive nitrogen in the atmosphere and must be treated as realistically as possible in global chemical transport models in order to obtain reasonable upper tropospheric budgets of NO_x and O₃. The global lightning parameterization of Allen and Pickering [2002] uses six-hour averaged upward cloud mass flux from the NASA Goddard Data Assimilation System (GEOS-DAS) to estimate flash rates. The parameterization was developed using primarily middle latitude ground-based flash data, but was tested using global observations from the OTD satellite instrument. The parameterization performs reasonably well in many parts of the world, but particular deficiencies were noted in the tropics. Therefore, we have analyzed ground-based lightning flash data from NASA's Brazil Lightning Detection Network centered in Rondonia, Brazil in relation to upward cloud mass flux, convective precipitation rate, and cloud top height data from a GEOS assimilation for the TRMM period. Modified functional relationships between observed flash rates and these variables were derived and applied for specific periods. Results were compared with LIS flash rates during wet and dry seasons.