%0 Dataset
%T Long-term daily composite precipitation dataset for the Loess Plateau  from 2015 to 2100
%J National Cryosphere Desert Data Center
%I National Cryosphere Desert Data Center(www.ncdc.ac.cn)
%U http://www.ncdc.ac.cn/portal/metadata/ce7e05d6-888c-4f0c-9551-4d4de8e9b42a
%W NCDC
%R 10.12072/ncdc.loess.db7330.2026
%A ZHANG Baoqing 
%K precipitation;long term;deep learning;high-resolution
%X Precipitation time series data serve as critical indicators for global and regional climate and hydrological cycle research. However, due to the lack of spatiotemporally continuous, high-quality time series datasets, significant uncertainties persist in current assessments of long-term precipitation dynamics and changes at global and regional scales. This study utilizes future precipitation datasets from ACCESS-ESM1-5, IPSL-CM6A-LR, and MIROC-ES2L, and constructs a downscaling model based on the U-Net deep learning architecture to generate a future precipitation data product for the Loess Plateau from 2015 to 2100. The spatial coverage of the dataset spans 33.75°N to 41.25°N and 100.95°E to 114.55°E, with a high spatial resolution of 0.1° and daily temporal resolution. Compared to the original model data, the core advantage of this dataset lies in its high resolution, which enables more precise characterization of precipitation details in complex terrain, significantly enhancing its practical value for regional hydrological modeling and extreme precipitation event analysis, among other research applications.The data are stored in NetCDF format (*.nc) with the file naming convention "LoessPlateauRegion_ACCESS-ESM1-5_ssp126_Prec_2015_2100_merged," where LoessPlateauRegion denotes the area name, ACCESS-ESM1-5 represents the model name, and ssp126 indicates the future scenario name. Within the data file, the variable "daily_precipitation" provides daily precipitation amounts at a daily temporal resolution and 10 km spatial resolution for the period 2015–2100, while "lat" and "lon" specify the latitude and longitude of the model grid points, respectively.