Work

Daily-scale dataset of highly dynamic water of Poyang Lake

Year: 2025

Type: article

Abstract: Poyang Lake experiences cyclical height changes annually, serving as a prime focus for highly dynamic water resource research. However, its complex environment and cloudy climate make timely and accurate mapping of water resources difficult, as well as more accurate ecological and water resource studies. We introduced the Tanh activation function and constructed a new optical water index, the Combined Water Index (CWI), to improve the accuracy and automation level of water identification. Meanwhile, in view of the cloudy weather conditions in Poyang Lake, we integrated the microwave images and proposed the New SAR Water Index (NSWI), which effectively supplemented the observation data. We established a 37-year-long observation experience of the water of Poyang Lake from 1986 to 2023. Third, we introduced water level, cumulative outflow from the Three Gorges Dam, and precipitation data to build a predictive model for random forests with reliable accuracy. Finally, we realized the spatial inversion of the predicted area and constructed a daily-scale water dataset of Poyang Lake (2016–2021). This study offers critical data support for advancing Poyang Lake ecological research and water resource management.

Source: Big Earth Data

DOI: 10.1080/20964471.2025.2506289