Probabilistic Site Adaptation for High-Accuracy Solar Radiation Datasets in the Western Sichuan Plateau

This dataset focuses on the high-precision quantification of Downward Shortwave Radiation (DSR) at the Earth’s surface over the Western Sichuan Plateau (WSP), aiming to provide reliable data support for regional solar energy resource assessment and development. The data generation process is based on ground-based observations as the benchmark, involving the correction and optimization of multi-source gridded DSR products using 12 Probabilistic Site Adaptation (PSA) methods: high-quality short-term in-situ data were first acquired by surface pyranometers, serving as unbiased reference standards to systematically correct the hourly DSR raw data from 2 satellite products and 2 reanalysis products; subsequent comparative evaluation of three methodological categories (benchmarking, parametric/non-parametric, and quantile combination approaches) led to the selection of the optimal Median Ensemble (MED) method, which achieved significant error reduction (RMSE: 163.97 W/m², nRMSE: 34.43%) and effectively corrected the inherent errors (RMSE > 200 W/m²) commonly present in the original gridded datasets. Techniques such as radiative transfer model calibration and statistical ensemble analysis were comprehensively applied during data processing to ensure data accuracy and consistency.The dataset covers the entire hourly observation period of 2018, with a spatial focus on the Western Sichuan Plateau region. It features a temporal resolution of 1 hour and a spatial resolution of 0.05° × 0.05°, fully recording the spatiotemporal distribution characteristics of solar radiation in the region throughout the year. The data are stored in 365 NetCDF (nc) files, with one independent file corresponding to each day, and there is no data missing, ensuring the continuity and integrity of the time series. Each nc file contains 8 hourly DSR grid data for the corresponding date; the core variable is the corrected surface downward shortwave radiation value, with a unit of measurement of W/m², accompanied by auxiliary information such as latitude, longitude, and timestamps to facilitate spatial analysis and temporal mining. The NetCDF format is a standard format widely used in the meteorological and geographical data fields, which can be read and processed using geographic information software such as GRASS GIS, QGIS, and ArcGIS, as well as programming tools including Python (with xarray and netCDF4 libraries) and R (with the ncdf4 package). The official download addresses of the relevant software are as follows: GRASS GIS (https://grass.osgeo.org/), QGIS (https://qgis.org/), and ArcGIS (https://www.esri.com/en-us/arcgis/products/arcgis-desktop/overview). Python and R can be configured with the corresponding libraries through their official websites. The methodological framework and data outcomes of this dataset provide practical technical references and data support for solar energy resource optimization in complex terrain regions.