A Methodology to Improve Satellite Quantitative Precipitation Estimates Using the UND Hybrid Classification Product and GOES Optical Depth Retrievals

Satellite Quantitative Precipitation Estimates (QPEs) from geostationary satellites have great strengths in spatial coverage and temporal resolution, but face significant limitations due to the indirect relationship between precipitation rates and brightness temperatures. Incorporation of optical depth into SCaMPR (Kuligowski 2002), a brightness temperature based precipitation estimation algorithm developed for use in the GOES-R era, produced significant improvements in rainfall estimates during convective events (Stenz et al. 2014, 2015). To identify the specific weaknesses of SCaMPR during deep convection, develop a conceptual model to address these weaknesses, and evaluate the applied modifications to SCaMPR, the UND Hybrid Classification Product (Feng et al. 2011) was used along with National Mosaic and Multi-Sensor (NMQ) Q2 precipitation estimates and OK MESONET observations. This combination of datasets allowed the evaluation and improvement of SCaMPR for individual Deep Convective System (DCS) components. An overview of the evaluation methodology from surface observations, to radar based precipitation estimates, and the UND Hybrid Classification Product will be presented, as this approach and the UND Hybrid Classification Product can be beneficial for improving numerous other satellite QPEs. Additionally, the improvements made to SCaMPR by incorporating optical depth retrievals will be discussed.