Seasonal Dynamics of the Land-Surface Characteristics in Arid Regions Retrieved by Optical and Microwave Satellite Data

Abstract

Establishing a quantitative relationship between Synthetic Aperture Radar (SAR) data and optical data can facilitate the fusion of these two data sources, enhancing the time-series monitoring capabilities for remote sensing of a land surface. In this study, we analyzed the Normalized Difference Vegetation Index (NDVI) and Shortwave Infrared Transformed Reflectance (STR) with the backscatter coefficients in vertical polarization VV (σ0VV) and cross polarization VH (σ0VH) across different seasons. We used optical and microwave satellite data spanning from the southern Gobi Desert region to the steppe region in northern Mongolia. The results indicate a relatively high correlation between the NDVI derived from Sentinel-2 and σ0VH (RVH = 0.29, RVH = 0.44, p 0VV (RVH = 0.06, RVH = 0.14, p 0VH and σ0VV except in spring, with the highest correlation coefficients observed in summer (RVV = 0.45, RVV = 0.44, p 0VH were noted, with a strong positive correlation peaking in summer (RVH = 0.71, p 0VV except in summer (RVV = −0.43, RVV = −0.34, RVV = −0.13, p 0VH and σ0VV in summer (RVH = 0.40, RVV = 0.39, p VH = 0.38, RVV = 0.09, p VH= −0.17, RVV= −0.38, p VH = −0.21, RVV = −0.06, p 0VH, and σ0VV were shown to vary by season and region. In the Gobi Desert region, perennial shrubs are not photosynthetic in spring and winter, and they affect backscatter due to surface roughness. In the steppe region, annual shrubs were found to be the dominant species and were found to photosynthesize in spring, but not enough to affect the backscatter due to surface roughness.

Keywords

• integration of satellite data and field data
• Sentinel-1 and Sentinel-2
• linear spectral unmixing
• Gobi Desert region
• steppe region
• Asian dust source area