Monitoring of subsidence water body change based on Google Earth Engine

With its economic benefits, coal mining also causes many environmental problems. In the mining areas with high phreatic water levels in the eastern China, the coal mining leads to surface subsidence, resulting in surface ponding. The drastic changes in the area of water bodies have reduced cultivated land and changed the agricultural ecosystem. Therefore, a continuous monitoring of coal mining subsidence is necessary for studying ecological changes in the region and formulating some restoration plans. In order to reveal the changes of water body in the coal mining subsidence area, based on the Google Earth Engine (GEE) platform, this study took the Panxie mining area as the study area and investigated the subsidence water data extracted from Landsat remote sensing image data from 1989 to 2016. The study constructed the ray method to obtain the interannual water boundary change distance, used the onevariable linear least squares regression method to fit the subsidence water boundary of the mining area, and analyzed the spatial position change of the subsided water body through the expansion coefficient. The results are as follows: ① The automatic recognition process of the direction change of water body in mining subsidence area by the ray method is proposed, and the method for selecting the ray origin and the angle between rays is clarified.② The study shows that the distance between the ray origin and the boundary of the subsidence water body has a strong correlation with its corresponding year through the Pearson correlation analysis of the constructed interannual ray distance change data. ③ The onevariable linear least square regression equation is constructed for each ray to predict the boundary of the subsidence water body, and the overall coefficient of determination of the fitting is 84.56%. ④ The degree of the fitting is good. The 2017 subsided water data predicted by the regression equation is compared with the 2017 subsided water data extracted from remote sensing images, and the prediction accuracy is 84.43%. ⑤ The expansion analysis shows that the subsidence water bodies of Xieqiao Mine, Pan No.3 Mine, Panbei Mine, Pan No.2 Mine, and Pan No.1 Mine expand slowly, while the subsidence water bodies of Zhangji Mine, Gubei Mine, Dingji Mine, and Zhuji Mine expand at a slower rate. The expansion speed of Guqiao Mine in the northwest direction is faster, and the expansion speed in the southeast direction is slow, which basically corresponds to the mining situation of mining enterprises. ⑥ Based on the GEE platform, from the extraction of subsidence water body to ray monitoring, the whole research is basically automated and can predict the subsidence ponding area quickly and accurately. The conclusion is that the subsidence water monitoring technology based on the ray method can reveal the changes of subsidence water in all directions to a certain extent, and provide a certain theoretical basis and data support for the ecological restoration of the mining area without coal mining related information.