Improvements to the AMSR-E Rain Over Land Facility Algorithm
Principal Investigator
Ralph R Ferraro
NOAA/NESDIS/ORA
Code E/RA2, Rm 601, WWBG Atmos Research & Appl Div, 5200 Auth Rd Camp Springs, MD 20746-4304
E-mail: ralph.r.ferraro@noaa.gov
Phone: 301-763-8251 x147
Fax: 301-763-8580
Abstract
The current AMSR-E level 2 rain algorithm continues to be generated through the Goddard Profiling Algorithm (GPROF). GPROF is currently at V6.5 for AMSR-E and the most recent improvements include better retrievals along coastlines, adjustments for the loss of one of the 89 GHz radiometers, and updated databases for the oceanic retrievals. Despite the progress in GPROF over the past several years, many in conjunction with the NASA Precipitation Measurements Mission (PMM) science team (with an emphasis on the TRMM 2A12 product), several improvements are urgently needed for the AMSR-E. These include:
- More realistic cloud hydrometeor profiles over land that capture a wider range of precipitation regimes (beyond the tropics)
- An overhaul of the land surface screening module which is outdated and should incorporate ancillary data sets, as well as other AMSR-E derived products (e.g., snow and sea-ice cover, soil moisture, emissivity) and additional channel measurements (e.g., 6.9 and 10.7 GHz).
- Inspection of region specific issues identified over the past several years by the user community (i.e., shallow convection, high terrain precipitation, etc.).
This project will be coordinated with a similar effort that will be proposed by the current members of the AMSR-E science team (e.g., C. Kummerow and T. Wilheit). The PI has led the development of the land portion of this algorithm. We will seek nominal funds to provide the necessary support to the precipitation over land portion of GPROF to advance its improvement specifically for AMSR-E, and to be able to address user concerns when they arise. Additionally, the resources are required to test new versions of the AMSR-E L2 rainfall products whenever reprocessing occurs (nominally, once each year) due to L1 data improvements (e.g., updated calibration) and L2 algorithm changes.
|