Key points are not available for this paper at this time.
Given the growing environmental challenges, accurate monitoring and prediction of changes in water bodies are essential for sustainable management and conservation. However, existing water extraction methods often rely on individual or cloud-free images, making them less robust for long-term, large-scale monitoring under varying environmental conditions. The Continuous Monitoring of Land Disturbance (COLD) algorithm provides a valuable tool for real-time analysis of land changes, such as deforestation, urban expansion, agricultural activities, and natural disasters. This study addresses the gap by leveraging COLD’s temporal modelling to estimate water frequency and detect water body changes. Our findings show that COLD-derived data can reliably estimate water frequency during stable periods and delineate water bodies with 0. 92 ± 0. 06 overall accuracy. Additionally, it enables the assessment of post-disturbance trends in water change, determining whether water frequency increases, decreases, or remains stable with 0. 81±0. 09 accuracy. The source code used in this study is publicly available at https: //github. com/hngpham/waterchangeCOLD/tree/published.
Pham et al. (Sat,) studied this question.