Batolar et al. (2026) Statistical Downscaling of Daily Mean Air Temperature of the High Asia Refined Analysis, Version2 (HARv2) in the North-West Himalaya (NWH), India

Identification

• Journal: Lecture notes in networks and systems
• Year: 2026
• Date: 2026-01-01
• Authors: Navdeep Batolar, Dan Singh, Mukesh Kumar
• DOI: 10.1007/978-981-95-2878-3_34

Research Groups

• Defence Geoinformatics Research Establishment (DGRE), Chandigarh, India
• University Institute of Engineering and Technology (UIET), Panjab University, Chandigarh, India

Short Summary

This study develops and evaluates three statistical downscaling methods (altitude correction, regression, and quantile-quantile mapping) to improve daily mean air temperature estimates from the High Asia Refined Analysis version 2.0 (HARv2) in the North-West Himalaya, India, demonstrating their effectiveness in generating long-length, localized temperature data.

Objective

• To develop, employ, and evaluate the performance of three statistical downscaling methods (altitude correction, regression, and quantile-quantile mapping) for daily mean air temperature from HARv2 reanalysis data at 10 stations in the North-West Himalaya, India.

Study Configuration

• Spatial Scale: 10 stations within the North-West Himalaya (NWH), India.
• Temporal Scale: Daily mean air temperature, with the aim of generating long-length data.

Methodology and Data

• Models used: Altitude Correction (ALTC), Regression (SR), Quantile–Quantile Mapping (SQ).
• Data sources: High Asia Refined Analysis version 2.0 (HARv2) reanalysis data and observed daily mean air temperature (OB) for validation.

Main Results

• The HARv2 daily mean air temperature showed a statistically significant positive correlation with observed daily mean air temperature at all stations.
• The Root Mean Square Error (RMSE) for the original HARv2 data in estimating observed daily mean air temperature ranged from 5.7–11.6 °C for training data and 6.3–11.9 °C for test data.
• After statistical downscaling, the RMSE for training data improved to 5.6–9.1 °C (ALTC), 4.2–7.8 °C (SR), and 4.2–7.8 °C (SQ).
• For test data, the downscaled RMSE improved to 6.3–8.6 °C (ALTC), 3.9–6.4 °C (SR), and 3.9–6.4 °C (SQ).
• Statistical downscaling significantly improved the estimation of observed daily mean air temperature in the NWH.
• The study demonstrated the feasibility of developing long-length daily mean air temperature data utilizing HARv2 for specific locations in the NWH.

Contributions

• Development and comparative evaluation of three statistical downscaling methods tailored for HARv2 daily mean air temperature in the complex, data-scarce terrain of the North-West Himalaya.
• Quantification of the significant improvement in accuracy of HARv2 temperature estimates achieved through the applied downscaling techniques.
• Provision of a robust methodology for generating long-length, localized daily mean air temperature datasets in high-altitude regions, which is critical for hydro-meteorological applications, climate change studies, disaster mitigation, and sustainable development planning.

Funding

• Not specified in the paper.

Citation

@article{Batolar2026Statistical,
  author = {Batolar, Navdeep and Singh, Dan and Kumar, Mukesh},
  title = {Statistical Downscaling of Daily Mean Air Temperature of the High Asia Refined Analysis, Version2 (HARv2) in the North-West Himalaya (NWH), India},
  journal = {Lecture notes in networks and systems},
  year = {2026},
  doi = {10.1007/978-981-95-2878-3_34},
  url = {https://doi.org/10.1007/978-981-95-2878-3_34}
}