Reduced major axis approach for correcting GPM/GMI radiometric biases to coincide with radiative transfer simulation. (January 2016)

• Record Type: Journal Article
• Authors: Islam, Tanvir; Srivastava, Prashant K.; Petropoulos, George P.; Singh, Sudhir K.
• Is Part Of: Journal of quantitative spectroscopy & radiative transfer. Volume 168(2016:Jan.)
• Page Start: 40
• Page End: 45
• Publication Date: 2016-01
• Subjects: Bias correction and estimation -- Radiometric adjustment -- Radiative transfer model -- Radiance assimilation -- Satellite calibration -- Global Precipitation Measurement (GPM); Spectrum analysis -- Periodicals; Radiation -- Periodicals; Analyse spectrale -- Périodiques; Rayonnement -- Périodiques; Radiation; Spectrum analysis; Periodicals; 543.0858
• DOI: 10.1016/j.jqsrt.2015.08.016
• Languages: English
• ISSNs: 0022-4073

Abstract

Abstract: Correcting radiometric biases is crucial prior to the use of satellite observations in a physically based retrieval or data assimilation system. This study proposes an algorithm – RARMA (R adiometricA djustment usingR educedM ajorA xis) for correcting the radiometric biases so that the observed radiances coincide with the simulation of a radiative transfer model. The RARMA algorithm is a static bias correction algorithm, which is developed using the reduced major axis (RMA) regression approach. NOAA's Community Radiative Transfer Model (CRTM) has been used as the basis of radiative transfer simulation for adjusting the observed radiometric biases. The algorithm is experimented and applied to the recently launched Global Precipitation Measurement (GPM) mission's GPM Microwave Imager (GMI). Experimental results demonstrate that radiometric biases are apparent in the GMI instrument. The RARMA algorithm has been able to correct such radiometric biases and a significant reduction of observation residuals is revealed while assessing the performance of the algorithm. The experiment is currently tested on clear scenes and over the ocean surface, where, surface emissivity is relatively easier to model, with the help of a microwave emissivity model (FASTEM-5). Highlights: Applicability of a new bias correction technique is examined. The experiment is conducted on the GPM Microwave Imager (GMI). The CRTM is used as the radiative transfer model. Proposed algorithm has shown significant reduction of observation residuals.