Monitoring Soil Salinity Using Machine Learning and the Polarimetric Scattering Features of PALSAR-2 Data

Jing Zhao,Ilyas Nurmemet,Nuerbiye Muhetaer,Sentian Xiao,Adilai Abulaiti

DOI: https://doi.org/10.3390/su15097452

IF: 3.9

2023-05-02

Sustainability

Abstract:Soil salinization is one of the major problems affecting arid regions, restricting the sustainable development of agriculture and ecological protection in the Kriya Oasis in Xinjiang, China. This study aims to capture the distribution of soil salinity with polarimetric parameters and various classification methods based on the Advanced Land Observing Satellite-2(ALOS-2) with the Phased Array Type L-Band Synthetic Aperture Radar-2 (PALSAR-2) and Landsat-8 OLI (OLI) images of the Keriya Oasis. Eleven polarization target decomposition methods were employed to extract the polarimetric scattering features. Furthermore, the features with the highest signal-to-noise ratio value were used and combined with the OLI optimal components to form a comprehensive dataset named OLI + PALSAR2. Next, two machine learning algorithms, Support Vector Machine (SVM) and Random Forest, were applied to classify the surface characteristics. The results showed that better outcomes were achieved with the SVM classifier for OLI + PALSAR2 data, with the overall accuracy, Kappa coefficient, and F1 scores being 91.57%, 0.89, and 0.94, respectively. The results indicate the potential of using PALSAR-2 data coupled with the classification in machine learning to monitor different degrees of soil salinity in the Keriya Oasis.

environmental sciences,environmental studies,green & sustainable science & technology