Blue Network - for SDAD Seminar Series

Social and Decision Analytics Seminar Series: Preparing Space-Based Observations for Comparisons with Climate Models

Event Details

Wednesday, October 5 2022
1:00pm-2:00pm Eastern Time (ET)


This event is part of our Social and Decision Analytics Seminar Series.

Jonathan Wynn Smith's Headshot

Speaker: Dr. Jonathan Wynn Smith, Meteorologist, National Oceanic and Atmospheric Administration

Abstract: A necessary step to future model development of lighting flash frequency (ff) and lightning - nitrogen oxide parameterizations is to compare space-based lightning observations to Earth System model ff output. This analysis compares ff’s in the novel GOES-16 Geostationary Lightning Mapper (GLM) to GFDL’s Atmospheric Model version 4 (AM4) model output for the years 2018-2020. This study counted the number of lightning flashes incident within a 2 or 4 km2 grid cell area and divided this by the number of seconds in a month. The GLM field of view is superimposed on the GOES-16 approximate 81.5°N/°S to 156°W-6°E grid. Which in 2018 contained a 4 km2 2711 x 2711 curvilinear grid and post 2018 a 2 km2 5424 x 5424 curvilinear grids. The AM4 model spatial resolution of 1° latitude by 1.25° longitude resulted in a destination grid resolution of 162 x 130 grid. The very fine observational grid downscaling to the model grid required conservative regridding. This technique averages the fine source full and partial grid values to the corresponding coarse destination grid. This required installing both the Earth System Modeling Python (ESMPy) and xesmf v8 libraries on GFDL supercomputing infrastructure. I used NCAR Common Language (ncl) to create the source and destination grid files. The initial comparisons suggest that the AM4 model underpredict by 2 orders of magnitude (1x10-5 to 1x10-7) in the Northern Hemisphere (NH) summer and 1 order of magnitude (1x10-5 to 1x10-6) in the Southern Hemisphere (SH). With the exception of overpredictions in the North American Rocky and South American Andes Mountains throughout the years of 2018-20 and a general poor correlation of less than 0.25 over the ocean. To improve the correlations and diminish biases, this study will compare the model output and GLM ff’s to polar-orbiting Lightning Imaging Sensor (LIS) on the TRMM-and International Space Station - LIS ff’s.

Join this event here.