H24C-02:
Evaluating the Global Precipitation Measurement Mission with NOAA/NSSL Multi-Radar Multisensor: Past, Current Status and Future Directions.
Tuesday, 16 December 2014: 4:15 PM
Pierre-Emmanuel Kirstetter1, Yang Hong1, Jonathan J Gourley2, Nicholas Carr3, Walter Arthur Petersen4, Mathew Schwaller5, Emmanouil N Anagnostou6, Christian D Kummerow7, Ralph R Ferraro8 and Nai-Yu Wang9, (1)University of Oklahoma, Norman, OK, United States, (2)National Severe Storms Lab, Oklahoma City, OK, United States, (3)University of Oklahoma Norman Campus, Norman, OK, United States, (4)NASA GSFC/WFF Code 610.W, Wallops Island, VA, United States, (5)NASA GSFC, Greenbelt, MD, United States, (6)Civil & Env Engineering, Storrs Mansfield, CT, United States, (7)Colorado State Univ, Fort Collins, CO, United States, (8)Univ Maryland-ESSIC/CICS and NOAA/NESDIS, College Park, MD, United States, (9)University of Maryland, College Park, MD, United States
Abstract:
Accurate characterization of uncertainties in precipitation estimates derived from space-borne measurements is critical for many applications including water budget studies or prediction of natural hazards caused by extreme rainfall events. GPM precipitation level II estimates are compared to the NEXRAD-based precipitation estimates derived from NOAA/NSSL’s Multi-Radar, Multisensor (MRMS) platform. The NEXRAD network has undergone an upgrade in technology with dual-polarization capabilities. These new polarimetric variables are being incorporated in MRMS to improve quality control of reflectivity data and to correct for partial beam blockages. The MRMS products, after having been adjusted by rain gauges and passing several quality controls and filtering procedures, are 1) accurate with known uncertainty bounds and 2) measured at a resolution below the pixel sizes of the GPM radar and radiometer observations. They are used by a number of NASA investigators to evaluate level II and level III satellite rainfall algorithms. The at-launch GPM Radiometer algorithm uses matches of coincident overpasses of various radiometers with surface rainfall from the MRMS database developed for the GPM project. Statistics from TRMM level II products serve as a benchmark to evaluate GPM precipitation estimates. Comparisons have been carried out at fine scale (e.g. instantaneous and 5 km for DPR) within a comparison framework developed to examine the consistency of the ground and space-based sensors in term of precipitation detection, characterization (e.g. convective, stratiform) and quantification. Specific error factors for passive (e.g. surface conditions for GMI) and active (e.g. attenuation of the radar signal, non uniform beam filling for DPR) sensors are investigated. Systematic biases and random errors quantified at the satellite estimation scale are useful for satellite-based Level III precipitation products. An online validation tool was designed to provide, for the first time, statistical evaluations and graphical representations of the uncertainties with GPM precipitation estimates over the CONUS. This cross-platform error characterization ultimately acts as a bridge to intercalibrate active and passive microwave measurements from the GPM core satellite to the constellation satellites.
