A New Extraction Method of Surface Water Based on Dense Time-Sequence Images

This is the authors' abstract. We don't add key points for this paper.

Fluctuations in the surface water are indicators of climatic and biological environmental variations. The water index method is the predominant approach for water extraction owing to its simplicity of operation and high efficiency. Recognizing the limitations of individual water indices in extracting water over dense time-sequences, this study introduces a combined water index (CWI) frequency method to improve the water extraction results. The research findings indicate the following: (1) CWI demonstrates superior extraction accuracy for various types of water when compared to other water indices, underscoring its higher precision and broader applicability. (2) By integrating CWI with the water frequency method, we propose an effective approach for dynamically monitoring water. This method accurately reflects changes in water under different conditions within dense time-sequence images. (3) Our results highlight the method's ability to precisely monitor dynamic water changes, efficiently extract various water types from Sentinel-2 data, and its potential for large-scale surface water mapping applications.