A 1km monthly precipitation dataset for the Qinghai-Tibetan Plateau using machine learning spatial interpolation

High-quality precipitation products with high spatial and temporal resolution play a key role in the study of the hydrological cycle at global and regional scales. Site-based interpolated datasets provide irreplaceable products for monitoring the spatial and temporal variability characteristics of precipitation, especially in areas with harsh natural conditions and lack of information. To this end, month-by-month precipitation data for the Tibetan Plateau were generated using station data from the National Meteorological Information Centre of the China Meteorological Administration (CMA) through a gated convolutional neural network with the introduction of an attention mechanism. The spatial resolution of this data is 0.0083333° (about 1 km), the temporal range is from January 1980 to December 2020, and the data format is TIF. Moreover, validation was carried out using data from independent meteorological observation sites, and the results proved that the dataset is credible. The unit of precipitation in the dataset is 1 mm. The TIF data can be opened for mapping by ArcMAP software , and it is recommended to use the WGS84 coordinate system.