Fullhart et al. (2025) Climate adaptation in the southwest US: The SWPar4.5 parameter set for stochastic weather generators

Identification

• Journal: Scientific Data
• Year: 2025
• Date: 2025-11-19
• Authors: Andrew T. Fullhart, Shang Gao, Wenting Wang, Emile Elias, Gerardo Armendariz, David C. Goodrich
• DOI: 10.1038/s41597-025-06102-5

Research Groups

• School of Natural Resources and the Environment, University of Arizona, Tucson, USA
• Department of Geographic Science, Beijing Normal University Zhuhai Campus, Zhuhai, China
• Southwest Climate Hub, USDA, Las Cruces, USA
• Ecoimpact Solutions, Durango, USA
• Southwest Watershed Research Center, USDA-ARS, Tucson, USA

Short Summary

This paper introduces Southwest Parameter Set 4.5 (SWPar4.5), a framework for generating probable historical and future point-scale climate time series at approximately 800 m resolution for the southwestern US, revealing increases in local precipitation intensity with implications for environmental indicators.

Objective

• To develop and provide the Southwest Parameter Set 4.5 (SWPar4.5) to enable a framework for creating probable historical and future point-scale climate time series at approximately 800 m resolution for targeted climate assessment in the southwestern US, addressing limitations of coarser gridded climate data.

Study Configuration

• Spatial Scale: Southwestern US (approximately 1.1 x 10^6 km² area comprising Nevada, Utah, Arizona, and New Mexico); point-scale data at approximately 800 m resolution (30 arc seconds).
• Temporal Scale: Historical period: 1974–2013 (40 years); Future periods: eight 30-year windows within 2000–2099 at decadal intervals. Daily and sub-daily time series generated.

Methodology and Data

• Models used:
  • Stochastic Weather Generators (SWGs), specifically CLIGEN.
  • Empirical Bayesian Kriging (EBK) for spatial interpolation.
  • Gradient Boosting (GB) regression for MX.5 P estimation.
  • Delta downscaling for precipitation parameters (MEANP, SDEVP, SKEWP).
• Data sources:
  • NEX-DCP30 (gridded climate projection, 30 arc sec, monthly, 1974–2099).
  • MACAv2 (gridded climate projection, 1/24 arc deg, daily, 1974–2099).
  • U.S. CLIGEN Ground Network (point-scale parameter sets, 1974–2013).
  • 3DEP 10 m DEM (elevation data, ~800 m resampled).
  • CMIP5 ensemble (specifically CCSM4, CanESM2, MIROC5 GCMs for RCP4.5 scenario).
  • NOAA-GHCNd (ground network for validation).

Main Results

• SWPar4.5 provides monthly climate benchmarks to parameterize stochastic weather generators, enabling the creation of probable historical and future point-scale climate time series at approximately 800 m resolution.
• The dataset reveals increases in precipitation intensity at the local scale, with monthly mean 30-minute maximum intensity (MX.5 P) increasing by approximately 10–16% for the 2070–2099 period, consistent with Clausius-Clapeyron scaling for a 2.7 °C regional temperature increase.
• Despite regional total annual accumulation remaining relatively stable, point-scale precipitation factors (standard deviation and skewness of daily non-zero accumulations) show upward trends, indicating precipitation becoming more "uneven" and skewed towards larger events.
• The framework allows for the calculation of various climate factors (e.g., annual rainfall erosivity, liquid water equivalent of snowfall) at high resolution, which are sensitive to sub-daily precipitation patterns and temperature changes.
• Comparison with other datasets shows SWPar4.5 achieves a good balance in predicting daily accumulation amounts, net accumulation, and the number of wet days, outperforming some gridded datasets in certain aspects.

Contributions

• Provides a novel, high-resolution (approximately 800 m) parameter set (SWPar4.5) for stochastic weather generators (specifically CLIGEN) for the southwestern US, addressing the "drizzle effect" and misrepresentation of local dynamics in coarser gridded climate data.
• Enables the generation of point-scale historical and future climate time series, including sub-daily precipitation patterns, crucial for site-specific environmental models (e.g., WEPP, RHEM).
• Offers a data fusion framework that minimizes steps and assumptions by directly using existing gridded climate projections and ground-based observations.
• Facilitates targeted climate assessment and adaptation planning by highlighting localized changes in precipitation intensity and other climate factors that are not apparent at regional scales or in coarser data.
• Provides tools (website and command-line executable) for easy querying and integration of the parameter set with CLIGEN.

Funding

• USDA, Agricultural Research Service, Rangeland Management Research Unit (CRIS Project # 3050-11210-007D).
• Southwest Climate Hub (CRIS Project # 3050-12610-001-000D).
• USDA Natural Resources Conservation Service (NRCS) Conservation Effects Assessment Project-Grazing Lands (CEAP-GL) (agreement NR213A750023C013).
• Long-Term Agroecosystem Research (LTAR) network.
• US Department of Agriculture, Agricultural Research Service (Agreement No. 58-2022-0-009).

Citation

@article{Fullhart2025Climate,
  author = {Fullhart, Andrew T. and Gao, Shang and Wang, Wenting and Elias, Emile and Armendariz, Gerardo and Goodrich, David C.},
  title = {Climate adaptation in the southwest US: The SWPar4.5 parameter set for stochastic weather generators},
  journal = {Scientific Data},
  year = {2025},
  doi = {10.1038/s41597-025-06102-5},
  url = {https://doi.org/10.1038/s41597-025-06102-5}
}