SRHrTD: A high-resolution air temperature dataset for the glacierized high mountain regions of Asia (1980-2022)

SRHrTD is a high-resolution near-surface air temperature (SAT) dataset developed for the high mountain regions of Central Asia over the period 1980–2022. The dataset provides monthly mean temperature fields at a spatial resolution of 1 km, enabling detailed investigations of long-term climate variability, elevation-dependent warming, cryospheric processes, and mountain hydrology in one of the world's most climate-sensitive regions. The dataset was generated through an integrated framework that combines pressure-level-based atmospheric downscaling with a machine-learning approach. Multiple sources of information, including in situ meteorological observations, ERA5 reanalysis data, topographic variables, and land-surface characteristics, were incorporated to reconstruct spatially continuous temperature fields over complex mountain terrain. Extensive evaluations against independent station observations demonstrate that SRHrTD provides improved performance relative to several widely used temperature datasets across Central Asia. The present release provides the complete monthly mean SAT dataset together with an example hourly NetCDF file (198001.tar), which contains hourly SAT fields for January 1980. The example file preserves the same data structure, metadata, coordinate system, and variable definitions as the complete hourly dataset, allowing users to understand the organization of the original hourly product. Because the complete hourly dataset (1980–2022) exceeds approximately 3 TB, only the example hourly file is included in the public archive. The complete hourly dataset is available from the corresponding author upon reasonable request. Spatial coverage: High Mountain Central Asia (HMCA) Temporal coverage: January 1980–December 2022 Spatial resolution: 1 km Monthly temporal resolution: Monthly Hourly example temporal resolution: Hourly (January 1980) Data format: NetCDF The monthly dataset is distributed as annual NetCDF files, with one file provided for each year from 1980 to 2022. Each file contains monthly mean near-surface air temperature fields for all 12 months of the corresponding year. The primary temperature variable is stored under the name monthlyMean, with dimensions of longitude, latitude, and month. Geographic coordinates are provided as lon and lat, and the month index (1–12) is stored in the variable month. The archive also includes an example hourly NetCDF file (198001.tar), which stores hourly near-surface air temperature for January 1980. The primary variable is 'temp', together with the coordinate variables lon, lat, and time, where time records hourly timestamps throughout the month. The monthly and hourly products share the same spatial grid, coordinate reference system, metadata, and variable naming conventions. The data are provided in NetCDF format and can be readily accessed using common scientific software packages that support the NetCDF standard, including MATLAB, Python, R, Panoply, and other compatible applications. The complete monthly dataset together with the example hourly file facilitates both direct scientific applications and the understanding of the organization of the original hourly SRHrTD product.