Poyang Lake has been frequently affected by droughts, which has drawn increasing attention. Floodplain lakes surrounding the main Poyang Lake play an important role in maintaining the hydrological cycle and supporting ecological diversity. Monitoring the dynamics of floodplain lakes is essential for guiding adaptive water resources management. However, lacking in-situ high temporal resolution data limits the detection of short-term water area variations in floodplain lakes. This study focused on the high-frequency hydrological changes of the sub-lakes within Poyang floodplain. A hierarchical spatiotemporal adaptive fusion mode (HSTAFM) was employed to generate NDVI imagery with an 8 d/30 m spatiotemporal resolution to obtain the water extent series of floodplain lakes. A total of 77 floodplain lakes (>1 km²) were selected to investigate their variations during 2000−2023. The nonlinear relationships of water areas of floodplain lakes with the water level of the main lake and the catchment inflows were further explored. Our findings indicated that: (1) the fused imagery was highly effective for water extraction for small lakes. The overall accuracy was 91% compared to the results from the Sentinel-2 imagery. (2) 94% of floodplain lakes had experienced an accelerating expansion during the rising phase, while 73% showing an accelerated shrinkage during the retreating phase. (3) A counterclockwise hysteresis between floodplain lakes areas and both main lake level and catchment inflows indicated a nonlinear response of floodplain lakes to these factors. This study extended our understanding of the hydrological processes of the floodplain lakes and provided methods to the studies in other floodplain lakes.
Jia et al. (2025) studied this question.