Abstract: Remote sensing technology brings new ideas for reservoir capacity curve reconstruction, but there are still problems such as the difficulty of direct application of multi-source remote sensing image data and low efficiency of reservoir batch area extraction, which make it difficult to form a systematic and process-oriented method for wide application. In view of this, this paper proposes a reservoir capacity curve reconstruction method coupled with multi-source time-phase remote sensing image data. The water level-volume relationship curve of Shenzhen Jingxin reservoir is reconstructed by this method. The experiments prove that the deviation rate of the reconstructed reservoir capacity curve is about 4.23%~11.5% in the range of normal water level, which can provide reference for the scheduling management of Jingxin reservoir. At the same time, it has reference significance for reservoirs in the country where the reservoir capacity curve cannot be reconstructed in time.

Key words: reservoir capacity curve reconstruction, multi-source image data coupling, reservoir area extraction, remote sensing, data fitting, reservoir computation models