Machine Learning Opportunity for Global Weather Radar

This is the first of a three-part series covering the innovative work by 557th Weather Wing for ongoing development efforts to leverage machine learning for a weather radar depiction across the globe, designated the Global Synthetic Weather Radar.
August 7, 2020
557th Weather Wing

“Radar data is extremely valuable environmental intelligence as it offers both operators and meteorologists insight into the state of the atmosphere,” said Chris Finnigsmier, 557th WW’s technical director. “Unfortunately, radar images are limited to areas adjacent to physical radar systems and thus unavailable across vast swaths of the planet.”

Research and development on the GSWR started when Massachusetts Institute of Technology’s Lincoln Laboratory combined weather data from existing sources and applied machine-learning techniques to create synthetic radar-like mosaics in areas just offshore of the continental United States with no radar coverage.   

Air Force leaders recognized the potential of these efforts, especially in high interest areas outside the continental U.S., and sponsored MIT/LL to produce global radar-like mosaics. The 557th WW’s high-resolution weather model, combined with satellite data from U.S. and allied sources, along with commercial global lightning data, feed the GSWR’s ability to conduct machine learning model training against actual precipitation data that has been collected by NASA.