Combining Machine Learning and Compact Polarimetry for Estimating Soil Moisture from C-Band SAR Data

Emanuele Santi,Simone Pettinato,Mohammed Dabboor,Simonetta Paloscia

doi:10.3390/rs11202451

Emanuele Santi, Simone Pettinato + Show 2 more

Open Access

https://doi.org/10.3390/rs11202451

Copy DOI

Abstract

This research aimed at exploiting the joint use of machine learning and polarimetry for improving the retrieval of surface soil moisture content (SMC) from synthetic aperture radar (SAR) acquisitions at C-band. The study was conducted on two agricultural areas in Canada, for which a series of RADARSAT-2 (RS2) images were available along with direct measurements of SMC from in situ stations. The analysis confirmed the sensitivity of RS2 backscattering (σ°) to SMC. The comparison of SMC with the compact polarimetry (CP) parameters, computed from the RS2 acquisitions by the CP data simulator, pointed out that some CP parameters had a sensitivity to SMC equal or better than σ°, with correlation coefficients up to R ≃ 0.4. Based on these results, the potential of machine learning (ML) for SMC retrieval was exploited by implementing and testing on the available data an artificial neural network (ANN) algorithm. The algorithm was implemented using several combinations of σ° and CP parameters. Validation results of the algorithm with in situ observations confirmed the promising capabilities of the ML techniques for SMC monitoring. Furthermore, results pointed out the potential of CP in improving the SMC retrieval accuracy, especially when used in combination with linearly polarized σ°. Depending on the considered input combination, the ANN algorithm was able to estimate SMC with Root Mean Square Error (RMSE) between 3% and 7% of SMC and R between 0.7 and 0.9.

Highlights

The capabilities of microwave satellite sensors, both active and passive, for observing the surface soil moisture content (SMC) in different environmental conditions and under various vegetation covers have been widely assessed in several research studies (e.g., References [1,2,3,4,5,6])
This study aims at exploiting the joint capabilities of compact polarimetry (CP) and artificial neural networks (ANN) for retrieving SMC in agricultural areas from fully polarimetric RADARSAT-2 (RS2) acquisitions: to our knowledge, while the retrieval of SMC by using full polarimetric RS2 acquisitions and ANN has been already attempted by other authors [30], CP and ANN have not been previously combined for the SMC retrieval in other studies
This study was aimed at demonstrating the joint use of machine learning (ML) and compact polarimetry (CP) for improving the retrieval of surface soil moisture (SMC) from synthetic aperture radar (SAR) acquisitions at C-band

Summary

Introduction

The capabilities of microwave satellite sensors, both active and passive, for observing the surface soil moisture content (SMC) in different environmental conditions and under various vegetation covers have been widely assessed in several research studies (e.g., References [1,2,3,4,5,6]). Several examples of SMC products generated by microwave satellite acquisitions are currently delivered by various space agencies through dedicated data portals (e.g., NASA NSIDC or JAXA GCOM). In this framework, synthetic aperture radar (SAR), was proven able to provide medium to high resolution mapping of SMC in all-weather conditions and independently of daylight. Research studies focusing on the use of Sentinel-1 (S-1) C-band SAR data, have been published with the aim of mapping temporal changes of surface SMC underneath agricultural crops. The validation of the algorithm is carried out by comparing the SMC estimated from the 1-km Sentinel-1 data over a central region (Umbria) in Italy and is validated against reference data with satisfactory results, considering the challenging topography and land cover of the test area

Objectives

Findings

Discussion

Conclusion