Monitoring method for water level in subsidence water bodies in mining areas using multi-source satellite observations

The high groundwater level mining area in the eastern part of China is an important mineral grain composite production area. The use of underground mining methods causes uneven subsidence on its surface, forming a large number of small and fragmented subsidence waterlogging areas. The dynamic changes of subsidence water level seriously affect the ecology and production safety of the mining area. Accurate monitoring of subsidence water level is crucial for wetland ecological management and flood risk prediction in high groundwater mining areas in the eastern region. Therefore, based on existing satellite altimetry technology, a method for monitoring the water level of fragmented subsidence water bodies in mining areas using multi-source satellite observations is proposed. Taking the Yanzhou mining area as the research object, we conduct an in-depth analysis of the correlation between the ICESat-2 (Ice, Cloud and Land Elevation Satellite-2) and GEDI (Global Ecosystem Dynamics Investigation) satellite observation datasets, and establishes a regression model. The model is used to correct the GEDI observation dataset, and by integrating multi-source satellite observation datasets, higher density and higher accuracy monitoring of subsidence water level in the mining area can be achieved. The results show that: ① The water level monitoring accuracy of ICESat-2 satellite altimeter is better, which can serve as the basis for evaluating and correcting GEDI data, and GEDI has a greater advantage in monitoring the quantity of water bodies; ② There is a significant correlation between ICESat-2 and GEDI observation datasets, and the insufficient number of surface cross observation points has a small impact on the fitting of the regression relationship. The use of regression models can effectively correct the water level observation data of GEDI satellites, reducing the average error, average absolute error, and root mean square error of water level measurement by 69.74%, 56.58%, and 50.56% respectively compared to before correction; ③ By integrating the ICESat-2 and GEDI observation datasets, high coverage and monitoring accuracy were demonstrated in water level monitoring, with a monitoring water coverage rate of 88.89% overall. The average error, average absolute error, and root mean square error of water level measurement were 0.10, 0.17, 0.26 m, respectively. Overall, the water level monitoring method for subsidence in mining areas based on multi-source satellite observations can meet the needs of small and scattered water level monitoring in mining areas, and also has potential applications in other problems that require high-density measurement.