An all-sky daily mean surface air temperature (Ta) product at 1 km spatial resolution over mainland China for 2003–2019 has been generated mainly from the Moderate Resolution Imaging Spectroradiometer (MODIS) products and the Global Land Data Assimilation System (GLDAS) dataset. Three Ta estimation models based on random forest were trained using ground measurements from 2384 stations for three different clear-sky and cloudy-sky conditions. The validation results showed that R2 and root mean square error (RMSE) values of the three models ranged from 0.984 to 0.986 and 1.342 K to 1.440 K, respectively, indicating that this high-resolution product has satisfactory accuracy. The format of the data is tif and a sample data is provided in example.zip. The dataset is stored by year, with 6 *.zip files per year. There is also an all-sky 0.01° daily surface air temperature product over Beijing for 2003–2019 (http://doi.org/10.5281/zenodo.4405123), which is a sub-dataset generated from this dataset for easy and convenient understanding of this dataset. And the sub-dataset has a data volume of only 264MB after compressed.

The dataset is also available at the University of Maryland (http://glass.umd.edu/Ta_China/). Due to the upload size restrictions, only the dataset for 2003–2008 is included on this site, and the complete dataset can be freely downloaded at http://glass.umd.edu/Ta_China/.