The need for the measurement response function (MRF) for optimal deconvolution of AMSR-E SSTs
Abstract
The current RSS AMSR-E SST products are produced on a 10×10km grid with each pixel representing SST averaged over a large area. This area is on the same order as the primary spectral channel used to determine SST (i.e., 75 km x 43 km) which is much bigger than the separation of adjacent samples ( 10km) and often approximated by a circle of diameter 56 km; i.e., the AMSR-E field is grossly oversampled. We are working on algorithms to deconvolve the oversampled AMSR-E fields. Success will us to produce a true 10 km, twice daily, global SST fields for the AMSR-E/AMSR2 period; an exciting prospect that is potentially of significant value in a broad range of important atmospheric and oceanic processes. Our initial approach was to use MODIS SST data both to aid in the deconvolution and to validate the results (AMSR-E and MODIS are on the same spacecraft). An issue, which has proven to be an impediment in this endeavor, is the lack of knowledge with regard to the weighting of SST values within the AMSR-E SST field, a requirement to perform an accurate deconvolution of this product. In light of this, the main objective of the work discussed herein is to estimate the measurement response function (MRF) associated with the AMSR-E SST product using matchups of MODIS SSTs. The MRF describes how much the SST from a particular distance and direction contribute to the observed AMSR-E value. In this study, we formulate the problem of the AMSR-E SST formation as a linear problem using a set of high-quality SST pixels and high-resolution co-located MODIS SST fields considered as the true SST fields. We then use the constrained least squares (LS) algorithm to estimate the MRF.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2018
- Bibcode:
- 2018AGUFM.C51D1099C
- Keywords:
-
- 3360 Remote sensing;
- ATMOSPHERIC PROCESSESDE: 0758 Remote sensing;
- CRYOSPHEREDE: 1855 Remote sensing;
- HYDROLOGYDE: 4275 Remote sensing and electromagnetic processes;
- OCEANOGRAPHY: GENERAL