Characterization of the Stepped Frequency Microwave Radiometer Calibration Errors and Their Impacts on Retrievals

The Stepped Frequency Microwave Radiometer (SFMR) is considered a facility instrument for hurricane reconnaissance and research. SFMR instruments in their current form have been flown routinely on the NOAA Hurricane Hunter P-3 aircraft and the Air Force Reserve 53^rd Weather Reconnaissance Squadron C-130J aircraft since 2004 and 2008, respectively. The SFMR wind speed and rain rate retrievals provide observational data to the National Hurricane Center forecasters, are being assimilated by weather models, and are widely distributed to the global operational and research communities. The SFMR data have also been extensively utilized to train tropical cyclone wind products from various satellite microwave remote sensing instruments. Recently an effort was started to characterize and document erroneous SFMR retrievals that have occasionally been observed in the inner core of hurricanes. Comparisons with low-level volume reflectivity and atmospheric wind retrievals from the Imaging Wind and Rain Profiler (IWRAP) identified several cases where the SFMR retrievals were clearly incorrect in regions of strong wind and rain gradients. It was assumed this was probably due mostly to the shortcomings in the geophysical model function utilized to infer wind speed and rain rate from the SFMR brightness temperature measurements. However, further investigation revealed significant calibration errors in the NOAA SFMR instruments that impact retrievals over the full range of wind speeds and rain rates. Upon further analysis of the SFMR calibration procedures it has become obvious that the existing calibration procedures are not adequate and that it is quite probable that all the SFMR instruments have varying degrees of calibration error that are adversely impacting their retrievals. The impacts on retrievals are not linear, which further complicates the problem. An analysis of these SFMR calibration errors, impacts on retrievals and a proposed path forward will be presented.