Intercomparison of Rain Rates from NOAA's Microwave Integrated Retrieval System (MiRS) with Surface Gauge Observations over China during 2014-2017

The Microwave Integrated Retrieval System (MiRS) is the official NOAA operational microwave-only satellite retrieval system, which was first introduced into operations in 2007. It currently processes passive microwave measurements from 10 different low earth orbit satellites, including ATMS data from the recently launched NOAA-20 satellite. The algorithm which is based on a one-dimensional variational approach uses the CRTM forward model to simulate radiances and to determine the atmospheric state which best fits the observed radiometric measurements, subject to other a priori constraints. Among the operational products produced are surface precipitation rate over both land and ocean.

MiRS precipitation (rain rate) retrievals from Suomi-National Polar Partnership (S-NPP) ATMS measurements have been extensively validated by way of regular daily comparisons with radar-rain gauge estimates over the U.S. and surrounding waters. However, validation of retrieved rain rates in other regions using surface gauges has been extremely limited. In this study, MiRS S-NPP/ATMS instantaneous rain rates were compared with surface rain gauge hourly measurements over China during the period 2014-2017, thus providing the first detailed quantitative assessment of MiRS operational rainfall retrievals for this region. Comparisons with rain gauge estimates were made each day for the period 2016-2017, providing an assessment of retrieval performance both regionally and seasonally. Additionally, assessment of rainfall was done for cases of 20 landfalling tropical cyclones (TCs) over China for the years 2014-2017 providing performance metrics specific to tropical cyclone conditions in Chinese coastal regions. Performance metrics will be presented that focus on probability distribution functions, categorical contingency scores, as well as statistical and spatial analyses. Additional analyses will assess inter-annual and regional differences, as well as differences due to changes in the operational retrieval algorithm.