Disaggregation of SMAP L3 Brightness Temperatures to 9km using Kernel Machines

TL;DR

This paper presents a machine learning approach using image segmentation and support vector regression to disaggregate SMAP brightness temperatures from 36km to 9km resolution, demonstrating high accuracy and potential for complex non-linear data.

Contribution

The study introduces a novel machine learning algorithm combining image segmentation and support vector machines for brightness temperature disaggregation at high resolution.

Findings

Disaggregated T$_{ extrm{B}}$ closely matches SMAP product with mean difference ≤ 5K.

Standard deviation of disaggregated T$_{ extrm{B}}$ is 7K lower than SMAP product.

Disaggregation method captures complex non-linear correlations effectively.

Abstract

In this study, a machine learning algorithm is used for disaggregation of SMAP brightness temperatures (T$_{\textrm{B}}$) from 36km to 9km. It uses image segmentation to cluster the study region based on meteorological and land cover similarity, followed by a support vector machine based regression that computes the value of the disaggregated T$_{\textrm{B}}$ at all pixels. High resolution remote sensing products such as land surface temperature, normalized difference vegetation index, enhanced vegetation index, precipitation, soil texture, and land-cover were used for disaggregation. The algorithm was implemented in Iowa, United States, from April to July 2015, and compared with the SMAP L3_SM_AP T$_{\textrm{B}}$ product at 9km. It was found that the disaggregated T$_{\textrm{B}}$ were very similar to the SMAP-T$_{\textrm{B}}$ product, even for vegetated areas with a mean difference $\leq$ 5K. However, the standard deviation of the disaggregation was lower by 7K than that of the AP product. The probability density functions of the disaggregated T$_{\textrm{B}}$ were similar to the SMAP-T$_{\textrm{B}}$. The results indicate that this algorithm may be used for disaggregating T$_{\textrm{B}}$ using complex non-linear correlations on a grid.