Inversion of multiwavelength radiometer measurements by three-dimensional filteringRemote sensing data from satellites typically have three dimensions: scan position, spacecraft position, and wavelength. Inversion of the radiometric data to infer geophysical parameters is a filtering problem in which the dimension of wavelength (or channel number) is transformed into a dimension of geophysical parameters, and the most general solution is a three-dimensional filter. Linear filters have the advantages of computational speed and easily described transfer functions; but often the measurements are nonlinear functions of the parameters to be inferred. To the extent that the nonlinear inversion problem is overdetermined, it can be modeled by a critically determined linear problem. As an example, inversion of Scanning Multichannel Microwave Radiometer (SMMR) data by means of a three-dimensional Wiener Filter is described. Atmospheric water vapor content, rain liquid water content, surface wind speed and surface temperature are the parameters inferred from the measurements. Nonprecipitating liquid water and water vapor scale height are also modeled but not retrieved. The a priori statistics on which the filter is trained have the effect of governing the selection of a trade-off point of noise as a function of resolution (in all three retrieval dimensions).
Document ID
19810061469
Acquisition Source
Legacy CDMS
Document Type
Conference Proceedings
Authors
Rosenkranz, P. W. (Massachusetts Inst. of Tech. Cambridge, MA, United States)
Baumann, W. T. (MIT Cambridge, MA, United States)
Date Acquired
August 11, 2013
Publication Date
January 1, 1980
Subject Category
Geophysics
Meeting Information
Meeting: Interactive Workshop on Interpretation of Remotely Sensed Data