Downscaled temperature and precipitation dataset for the Qilian Mountains region (1961–2022)

On the basis of considering terrain information, the Geographically Weighted Regression (GWR) method was used to downscale the coarse resolution (0.25 °) CN05.1 dataset. We obtained a monthly scale 1km dataset of temperature (Ta) and precipitation (Pre) in the Qilian Mountains from 1961 to 2022, and evaluated the downscaled 1km resolution dataset using observational data from meteorological stations. Compared with the original CN05.1 product, the root mean square error and standard deviation of the new dataset have a higher degree of agreement with meteorological station observations, which can provide detailed information for analyzing climate change trends at multiple time scales. This dataset will help potential users improve climate monitoring, modeling, and environmental research in the Qilian Mountains region. Dataset Description: (1) Time and spatial information of the dataset Time:January 1961 December 2022 92.09 ° -105.81 ° E, 34.57 ° -41.14 ° N. (2) Dataset Naming Ta1961-2021month.nc Pre1961-2021month.nc (3) Attribute information The data storage adopts NetCDF format. Each file contains 744 months of data and requires a total of 5.44GB of storage space. Each file name represents the data contained in the file. For example, the file named Ta1961-2022month.nc contains monthly temperature data from 1961 to 2022. The total number of NetCDF files is 2, and the total size of the nc format dataset is approximately 4.29 GB. The unit of temperature is ℃, and the unit of precipitation is mm. The spatial resolution of the data is 1km.Data processing methods: (1) Processing steps Firstly, prepare environmental variables with resolutions of 0.5 '(1 kilometer) and 25' and the original monthly scale CN05.1 data. Secondly, variables related to temperature and precipitation were selected and named as explanatory variables. Thirdly, input the explanatory variables and the original CN05.1 data into the GWR model to obtain intercepts, residuals, and coefficients. Fourthly, interpolate the intercept, residuals, and coefficients to obtain a 1km grid layer. Then combine these layers with explanatory variables at a 1-kilometer resolution to generate high-resolution temperature and precipitation data through GIS.(2) GWR model The GWR model is a regional regression model first proposed by Chris et al. (1996). This model is widely used for studying the dynamic characteristics and scale dependence of the relationship between dependent and explanatory variables.(3) GWR 4.0 software This dataset uses GWR 4.0 software（ http://geoinformatics.wp.st-andrews.ac.uk/gwr ）to establish GWR model.