Precipitation Retrievals Employing GOES Imager Infrared Channels and AMSU MIT Precipitation Retrieval Products

Abstract

This paper develops a neural network based precipitation retrieval algorithm called the PSU Infrared Precipitation retrieval algorithm version 1 (PIP-1), which employs infrared observations from the Geostationary Operational Environmental Satellite (GOES)-12 Imager. Neural networks are trained and evaluated using overlapping surface precipitation rates from the AMSU MIT Precipitation retrieval (AMP) products accurately retrieved using observations from the passive millimeter-wave spectrometer Advanced Microwave Sounding Unit (AMSU) aboard the U.S. National Oceanic and Atmospheric Administration (NOAA)-18 satellite. Results show good agreement between PIP-1 retrievals and AMP surface precipitation rates in terms of rates, positions, and morphology. PIP-1 retrievals are useful at rates higher than 1 mm/h. Employing PIP-1 with observations from GOES Imagers can provide useful surface precipitation retrievals in real time at every half an hour. PIP-1 can be adapted to work with other geostationary infrared satellites with similar channel characteristics to those of GOES Imagers.