Quantitative estimation of soil organic matter content in black soil region under varying climatic conditions using optical radar remote sensing technology

Using Heshan Farm, a representative black soil region, as a case study, this study explores the variability in optical and radar remote sensing signals in response to extreme variations in hydrothermal regimes across different climatic years. A multisource integration framework combining optimal bare-soil period imagery, multitemporal growing-season data, and stable environmental covariates was developed to improve the accuracy and stability of SOM retrieval. Using optical and radar remote sensing data from 2019–2021, three strategies were compared: bare-soil imagery, multitemporal fusion, and the integration of environmental covariates. The results showed that the optimal observation period differed by year, with the highest accuracy achieved in June 2020 (R2 = 0.660, RMSE = 6.003 g/kg). The integration of growing-season imagery (primarily September) improved the performance with R2 increasing by 0.030 and RMSE decreasing by 0.286 g/kg. Further inclusion of topographic, climatic, and soil physicochemical variables yielded maximum accuracy (R2 = 0.708, RMSE = 5.577 g/kg). Spatial analysis revealed elevated SOM concentrations in low-lying southern areas, whereas uncertainty analysis identified 2021 as the most stable prediction year (average uncertainty = 2.67 g/kg). Overall, the integration of multitemporal remote sensing data with environmental covariates substantially improved the precision and robustness of SOM spatial prediction.