HR: 16:15h
AN: H14B-02    [Abstracts]
TI: Calibration Corrections for Studying Land Hydrological Climate Trends Using Passive Microwave Satellite Observations
AU: * Chan, S K
EM: Steven.K.Chan@jpl.nasa.gov
AF: Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Drive, Pasadena, CA 91109 United States
AU: Njoku, E G
EM: Eni.G.Njoku@jpl.nasa.gov
AF: Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Drive, Pasadena, CA 91109 United States
AB: In climatological trend analysis, one often faces the task of building long-term measurement archives composed of remote-sensing data collected by a series of different spaceborne sensors over an extended period of time. Because of the hardware changes and advancements over time, systematic mismatches in calibrations often occur between different generations of instruments. These mismatches must be accounted for before a given uncorrected dataset can be properly interpreted, either individually or collectively with other calibrated datasets. In this study, we describe an empirical technique for performing such a calibration correction as the first step in a land climatology trend analysis. The technique makes use of a recent, relatively well-calibrated dataset to correct an earlier dataset with known calibration problems. Two three-year datasets, one from the Advanced Microwave Scanning Radiometer - EOS (AMSR-E, 2002-2005) and the other from the Scanning Multichannel Microwave Radiometer (SMMR, 1980-1983), were considered in this investigation. It was found that for a given frequency and polarization, a corrected brightness temperature (T_B) channel can be modeled as a weighted sum of the uncorrected H- and V-polarized T_B's plus a constant term, with regression coefficients determined from an optimal mixture of land and ocean observations characterized by low temporal variations over multiple annual cycles. With these observations serving as common reference T_B 'tie-points' for SMMR and AMSR-E, the proposed technique results in corrections that capture primarily the difference in measurement characteristics of the two sensors in response to the same physical T_B's. It was determined that this correction procedure results in regression residuals with bias less than 0.61 K in magnitude and standard deviation less than 3.41 K for the first three H- and V-polarized SMMR channels (6.63 GHz, 10.69 GHz, and 18.00 GHz) over a large area of such diverse surface classes as ocean, barren land, woody savannas, and evergreen forest.
DE: 1833 Hydroclimatology
DE: 1846 Model calibration (3333)
DE: 1872 Time series analysis (3270, 4277, 4475)
SC: Hydrology [H]
MN: Fall Meeting 2005