Spatial Distribution of Precipitation in Ji'an Based on Multi-source Data and GBDT Regression Model
In this study,the GBDT algorithm combined with multi-source remote sensing data sets are used to predict the precipitation in Ji'an.A precipitation forecasting model based on GBDT was constructed by integrating multiple sources such as DEM and ground station data to realize precipitation forecasting mapping in Ji'an.The results show that the GBDT model can comprehensively consider the information difference and correlation among different data sources,and integrate multi-source remote sensing data that can effectively reflect the fine distribution of precipitation.The predicted precipitation in Ji'an is 1 705-1 862 mm,showing a zonal distribution from southeast to northwest.The study provides important data support for water resources management and climate change research in Ji'an.