Dynamics of forest vegetation in East Asia over the past 40 years and its response to hydrothermal conditions

• Forest vegetation in East Asia increased 0.16% annually from 1982 to 2022. • Near-surface air temperature is the dominant driver of vegetation dynamics. • Temperature–moisture synergy strongly outperforms single hydrothermal factors • CNN-BiLSTM predicts continued NDVI increase by 2027 in East Asia. East Asian forest vegetation, central to regional ecosystems, is highly sensitive to climate change and varies with hydrothermal conditions. Based on NDVI and hydro-thermal variables (surface temperature, near-surface temperature, soil moisture, precipitation, and atmospheric water vapor) from 1982 to 2022, this study developed an integrated analytical framework that combines Seasonal-Trend decomposition using Loess(STL) for trend and seasonality extraction, Pruned Exact Linear Time(PELT) for detecting structural breakpoints, a partition-optimized geographical detector model for driver diagnosis and interaction detection, and a Convolutional Neural Network combined with Bidirectional Long Short-Term Memory (CNN-BiLSTM) model for future prediction. Results show that East Asian forest greenness increased at an average rate of 0.16% per year, with acceleration after 2010 and the fastest growth in the Yunnan-Guizhou Plateau and South China, whereas South Korea has exhibited degradation since 2017. Near-surface temperature emerged as the dominant driver, with its interactions with precipitation and soil moisture exerting the strongest influence on seasonal vegetation patterns. The CNN-BiLSTM model forecasts a continued greening, with an expected increase of 0.012 NDVI units by 2027. By integrating these complementary methods, this study clarifies the complex hydrothermal mechanisms driving forest vegetation dynamics and provides scientific support for regional ecological conservation and climate adaptation.