Improved Method Enhances Soil Moisture Mapping in Permafrost Regions
Feb 28, 2024
Scientists from the Aerospace Information Research Institute (AIR) with the Chinese Academy of Sciences (CAS) have developed an improved change detection method, based on the time series data from Sentinel-1 radar and Sentinel-2 optical sensors spanning the years 2019 to 2021, to estimate surface soil moisture. The approach sheds light on the intricate dynamics of the hydrological cycle and ecological environment in permafrost regions.
The study was published in the GIScience & Remote Sensing on Feb.5, 2024, and the corresponding dataset was published on the website of the National Tibetan Plateau Data Center.
In their study, the scientists expressed the response of backscatter to soil moisture in bare soil using a logarithmic form, departing from the traditional linear approach. Additionally, to account for the impact of vegetation on backscatter, an influence function of the normalized difference vegetation index (NDVI) was established for vegetation-covered surfaces. This effectively mitigated the influence of vegetation, allowing for accurate determination of the change in backscatter relative to bare soil conditions.
The researchers formulated an empirical function to ascertain reference (minimum and maximum) values of soil moisture in each pixel, facilitating accurate retrieval of soil moisture. The study focused on the Wudaoliang permafrost region of the Qinghai-Tibet Plateau, and the results were validated against ground measurements.
The findings demonstrated significant success, with correlation coefficients ranging from 0.672 to 0.941 and root mean squared errors (RMSE) ranging from 0.031 m3/m3 to 0.073 m3/m3 when comparing with individual in-situ measurements. Comparisons with the widely used Soil Moisture Active Passive (SMAP) 9-km product showed that the new method outperformed, boasting higher correlation (0.898 vs. 0.867) and lower RMSE (0.037 m3/m3 vs. 0.044 m3/m3) for capturing regional-scale soil moisture.
Furthermore, the soil moisture retrieved from Sentinel exhibited a strong correlation with the SMAP 9-km soil moisture over time, providing a more accurate representation of the region's soil moisture heterogeneity. This method showcases the feasibility of combining Sentinel-1 and Sentinel-2 for high-resolution (100 m) soil moisture mapping in permafrost regions.
For further information, please contact ZHAO Tianjie at zhaotj@aircas.ac.cn
Scientists and students set up ground observation instruments and collect soil samples in permafrost regions over Qinghai-Tibet Plateau. (Image by AIR)
News & Events