Modern Earth-observing geostationary satellites provide frequent high-resolution infrared radiance (GeoIR) observations of the atmosphere over much of the globe. The data assimilation (a type of Geospatial Big Data Fusion) of GeoIR observations thus have the potential to create high-quality high-resolution 4-dimensional reanalysis data of mesoscale convective systems (MCSs; a type of cloud system). However, existing reanalysis data for tropical MCSs (TMCSs) only 1) assimilate ~0.1% of all GeoIR observations and 2) do not possess the horizontal resolution needed to explicitly resolve the mesoscale structures of TMCSs. The creation of a high-resolution TMCS-resolving reanalysis dataset that better utilizes GeoIR observations thus have the potential to accelerate research into these socioeconomically important systems.

This talk presents the creation and validation of a high-resolution Tropical Mesoscale Convective Systems Reanalysis (TMeCSR; “tea-mixer”) community dataset. Every hour of TMeCSR data is created by assimilating more than 55,000 GeoIR observations (and a variety of other observations) into an ensemble of TMCS-resolving Weather Research and Forecasting (WRF) simulations. This dataset is publicly available for every hour of June, July, and August 2017, and spans the Indian Ocean, the Pacific warm pool, tropical continental Asia, and the Western Pacific. More than 1,200 TMCS events are captured by the TMeCSR. Most interestingly, the TMeCSR outperforms the gold standard ECMWF Reanalysis version 5 (ERA5) at resolving TMCSs and the diurnal rainfall cycle of the equatorial Maritime Continent. The TMeCSR is thus a veritable trove of publicly available and observation constrained gridded TMCS data that will likely benefit TMCS research for years to come.