Niu et al. (2026) AI-Refined Ensemble Precipitation and Temperature Projections for Regional Impact Assessment over the Conterminous United States

Identification

• Journal: Scientific Data
• Year: 2026
• Date: 2026-09-09
• Authors: Haoran Niu, Deeksha Rastogi, Shih‐Chieh Kao, Moetasim Ashfaq
• DOI: 10.1038/s41597-026-08043-z

Research Groups

• Computational Sciences and Engineering Division, Oak Ridge National Laboratory, United States
• Environmental Science Division, Oak Ridge National Laboratory, United States

Short Summary

This study presents an AI-refined ensemble of Earth system model (ESM) projections for the conterminous United States at 1/24° (~4 km) spatial resolution. The dataset was generated by downscaling ten CMIP6 ESMs under two emissions scenarios (SSP245 and SSP585) for an 80-year period (1980–2059).

Objective

• To provide high-resolution, long-term climate projections that can support regional hydroclimate analysis, climate-impact assessment, and related applications.

Study Configuration

• Spatial Scale: The dataset spans the geographic extent of 125°W–66.5°W and 24°N–53°N at a spatial resolution of 1/24° (~4 km).
• Temporal Scale: The dataset covers an 80-year period (1980–2059) under two emissions scenarios (SSP245 and SSP585).

Methodology and Data

• Models used: Super-Resolution Convolutional Neural Networks (SRCNN) and Super-Resolution Generative Adversarial Networks (SRGAN).
• Data sources: CMIP6 Earth system model outputs, Daymet observations.

Main Results

• The dataset provides daily high-resolution historical and future climate projections for precipitation, maximum temperature, and minimum temperature.
• Technical validation compares the generated products with Daymet observations and existing downscaled datasets to characterize spatial patterns, biases, temporal consistency, and differences among products.

Contributions

• This study presents a comprehensive evaluation of AI-based downscaling techniques for generating high-resolution climate projections, providing critical insights into hydrologic variability, future trends, and uncertainties.
• The dataset supports regional hydroclimate analysis, climate-impact assessment, and related applications by offering high-resolution, long-term climate projections.

Funding

• This research was supported by the US Department of Energy (DOE) under contract DE-AC05-00OR22725 with UT-Battelle, LLC.

Citation

@article{Niu2026AIRefined,
  author = {Niu, Haoran and Rastogi, Deeksha and Kao, Shih‐Chieh and Ashfaq, Moetasim},
  title = {AI-Refined Ensemble Precipitation and Temperature Projections for Regional Impact Assessment over the Conterminous United States},
  journal = {Scientific Data},
  year = {2026},
  doi = {10.1038/s41597-026-08043-z},
  url = {https://doi.org/10.1038/s41597-026-08043-z}
}