Ortiz et al. (2026) Spatiotemporal Distribution in Rainfall and Temperature from CMIP6 Models: A Downscaling and Correction Study in a Semi-Arid Region of Mexico

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Water
• Year: 2026
• Date: 2026-04-06
• Authors: Ricardo Robles Ortiz, Julián González Trinidad, Carlos Bautista Capetillo, Hugo Enrique Júnez Ferreira, Cruz Octavio Robles Rovelo, Ana Isabel Veyna Gómez, Sandra Dávila Hernández, Misael Del Rio Torres
• DOI: 10.3390/w18070874

Research Groups

[Not specified in the provided text.]

Short Summary

This study evaluated 15 CMIP6 models over Zacatecas, Mexico, and produced a 1 km historical climate dataset for 1985–2014 by statistically downscaling bias-corrected daily fields, assessing its performance against independent observational networks.

Objective

• To evaluate the performance of 15 CMIP6 models in resolving seasonal timing and topographic gradients over Zacatecas, Mexico.
• To produce a 1 km historical climate dataset for 1985–2014 for Zacatecas, Mexico, by statistically refining bias-corrected daily fields from NEX-GDDP-CMIP6.
• To assess the accuracy of the downscaled temperature and rainfall fields against independent observational baselines.

Study Configuration

• Spatial Scale: 1 km resolution, covering the region of Zacatecas, Mexico.
• Temporal Scale: Historical dataset for 1985–2014 (daily fields). Observational baselines cover 2004–2014 (INIFAP network) and 1985–2014 (CONAGUA network).

Methodology and Data

• Models used: 15 CMIP6 models (e.g., BCC-CSM2-MR, ACCESS-ESM1-5). Statistical downscaling techniques included an elevation-informed hybrid spline approach for temperature and geographically weighted regression (GWR) for rainfall.
• Data sources: NEX-GDDP-CMIP6 (bias-corrected daily fields), CHELSA climatology (downscaling reference), INIFAP automated network (observational baseline), CONAGUA conventional network (observational baseline).

Main Results

• For temperature, the BCC-CSM2-MR model demonstrated the highest performance, achieving a Pearson correlation coefficient (R) of 0.94 for both maximum and minimum daily temperatures.
• A consistent network-dependent bias pattern was identified for the diurnal temperature range: downscaled models overestimated it relative to the INIFAP network but underestimated it relative to the CONAGUA network.
• For rainfall, the ACCESS-ESM1-5 model effectively reproduced the seasonal cycle and dominant orographic patterns, with a correlation coefficient (R) of 0.611.
• The study successfully generated a 1 km spatially consistent historical climate dataset for Zacatecas, Mexico, spanning 1985–2014.

Contributions

• Development of a high-resolution (1 km) historical climate dataset for a complex semi-arid region (Zacatecas, Mexico), providing a valuable baseline for regional applications such as stochastic weather generation and impact models.
• Comprehensive evaluation of 15 CMIP6 models in a semi-arid context, identifying best-performing models for temperature (BCC-CSM2-MR) and rainfall (ACCESS-ESM1-5).
• Highlighting the critical influence of instrumentation and observational protocols on climate model evaluation through the identification of network-dependent bias patterns in diurnal temperature range.

Funding

[Not specified in the provided text.]

Citation

@article{Ortiz2026Spatiotemporal,
  author = {Ortiz, Ricardo Robles and Trinidad, Julián González and Capetillo, Carlos Bautista and Ferreira, Hugo Enrique Júnez and Rovelo, Cruz Octavio Robles and Gómez, Ana Isabel Veyna and Hernández, Sandra Dávila and Torres, Misael Del Rio},
  title = {Spatiotemporal Distribution in Rainfall and Temperature from CMIP6 Models: A Downscaling and Correction Study in a Semi-Arid Region of Mexico},
  journal = {Water},
  year = {2026},
  doi = {10.3390/w18070874},
  url = {https://doi.org/10.3390/w18070874}
}