The original data of the water area, water level, meteorological factors and sediment in Poyang Lake area.

This study investigates the spatiotemporal dynamics and multi-scale driving mechanisms of water area changes in Poyang Lake, China's largest freshwater floodplain lake, over 2000–2022. By integrating multi-source remote sensing (MODIS, Landsat), hydrological data from six stations (water level, sediment discharge), and meteorological records from five stations (precipitation, temperature, potential evapotranspiration, relative humidity), we applied Empirical Mode Decomposition (EMD) and Mann-Kendall mutation tests to decompose time series into seasonal, interannual, decadal, and long-term decadal components. Results reveal a significant shrinking trend of 28.02 km²/yr, with a critical regime shift around 2011. Driving factors exhibit strong scale-dependence: precipitation dominates seasonally; potential evapotranspiration becomes primary at interannual scale (variance contribution rate 25.64%, R²=0.91); temperature correlates with long-term interannual changes; and relative humidity drives long-term decadal variations. Spatially, the representativeness of hydrological stations shifts with scale—Kangshan (seasonal/long-term interannual), Duchang (interannual), Tangyin (decadal), and Hukou (long-term decadal) emerge as key stations, reflecting differential influences of upstream processes versus Yangtze River interactions. Sediment discharge at Hukou declined dramatically (144.1% over 23 years), serving as a proxy for human activities (sand mining, reservoir operations) that alter river-lake water-sediment exchange and exacerbate lake shrinkage. This study provides a transferable multi-scale analytical framework for understanding floodplain lake responses to climatic and anthropogenic pressures.本研究聚焦中国最大淡水洪泛平原湖——鄱阳湖，系统探究其2000—2022年间水域面积变化的时空动态特征与多尺度驱动机制。本研究整合多源遥感数据（MODIS、Landsat）、6个水文站点的观测数据（含水位、输沙量）以及5个气象站点的观测记录（含降水量、气温、潜在蒸散量、相对湿度），采用经验模态分解（Empirical Mode Decomposition，EMD）与曼-肯德尔突变检验（Mann-Kendall Mutation Test）方法，将时间序列拆解为季节尺度、年际尺度、年代际尺度与长期年代际分量。研究结果显示，鄱阳湖水域面积以28.02 km²/年的速率呈现显著萎缩趋势，并在2011年前后发生关键性状态跃迁。驱动因素表现出强烈的尺度依赖性：降水主导季节尺度的水域变化；潜在蒸散量在年际尺度成为首要驱动因子，其方差贡献率达25.64%，决定系数R²=0.91；气温与长期年际变化存在显著相关性；相对湿度则驱动长期年代际尺度的变异。空间维度上，水文站点的表征能力随尺度发生迁移：康山站适配季节与长期年际尺度、都昌站适配年际尺度、棠荫站适配年代际尺度、湖口站适配长期年代际尺度，成为关键观测站点，这反映了上游水文过程与长江交互作用的差异化影响。湖口站的输沙量在23年间大幅降低，降幅达144.1%，可作为人类活动（采砂、水库调度）的代理指标——此类人类活动改变了江湖水沙交换过程，进一步加剧了湖泊萎缩。本研究构建了可迁移的多尺度分析框架，为理解洪泛平原湖泊对气候与人为压力的响应机制提供了重要支撑。