All-sky Daily Surface Air Temperature over Mainland China at 1 km Resolution

An all-sky daily mean surface air temperature product at 1 km spatial resolution over mainland China for 2003–2019 was generated by Yan Chen of Wuhan University. The dataset was created mainly from MODIS and GLDAS data using random forest models trained on ground measurements from 2384 stations. Validation showed R² values from 0.984 to 0.986 and RMSE values from 1.342 K to 1.440 K, indicating satisfactory accuracy.

Use Cases

• Analyzing long-term temperature trends across mainland China based on the 2003-2019 daily time series.
• Validating climate models using the high-resolution (1 km) gridded temperature data.
• Studying urban heat island effects by comparing the all-sky temperature product with local station data.
• Assessing temperature variability under different sky conditions (clear-sky and cloudy-sky) using the three trained models.

Strengths

• High spatial resolution of 1 km for detailed regional analysis.
• Models were trained on ground measurements from 2384 stations, providing a substantial training base.
• Validation metrics indicate high accuracy, with R² values ranging from 0.984 to 0.986.
• Covers a 17-year period (2003-2019) for longitudinal climate studies.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Data is stored in yearly .zip files with 6 files per year, which may complicate initial data assembly.
• The description does not specify the exact file formats or data volume for the main dataset.

Provenance

Source

Yan Chen, Wuhan University

Collection Method

Generated from MODIS products and the GLDAS dataset using random forest models trained on ground station measurements.

Time Range

2003–2019

Geography

Mainland China