Kemsley et al. (2026) Mortality under locally unprecedented heat is better estimated with cross-regional learning

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

Identification

• Journal: Environmental Research Letters
• Year: 2026
• Date: 2026-06-26
• Authors: Sarah Wilson Kemsley, Jowan L. Fromentin, Bikem Pastine, Xiaowen Dong, Tom Matthews, Pierre Masselot, Katrin J. Meißner, Sarah E Perkins-Kirkpatrick, Louise J. Slater
• DOI: 10.1088/1748-9326/ae8317

Research Groups

• European Centre for Medium-Range Weather Forecasts (ECMWF)
• University of [Unknown]

Short Summary

This study demonstrates that jointly learning temperature-mortality relationships across regions in Europe improves estimation of excess mortality under locally unprecedented heat events and enhances predictive performance.

Objective

• Investigate the effectiveness of joint learning in improving temperature-mortality models, particularly for extreme heat events in regions with historically rare exposure.

Study Configuration

• Spatial Scale: Continental Europe (with a focus on heterogeneous heat exposure)
• Temporal Scale: Long-term observational data ( likely spanning multiple decades)

Methodology and Data

• Models used: Not specified ( likely machine learning or statistical models)
• Data sources: Satellite, observation, reanalysis data from various European regions

Main Results

• Joint learning improves estimation of excess mortality under locally unprecedented heat events.
• Largest gains in cooler climates with historically rare heat exposure (mean increase in pseudo-R2 = 0.09).
• Learned relationships remain epidemiologically interpretable and stable when incorporating concurrent environmental exposures.

Contributions

• This study contributes to the development of early warning systems for heat-related mortality by improving predictive performance, particularly in regions with heterogeneous heat exposure.
• The joint learning approach enables interpolation across a broader range of observed exposures, reducing reliance on local extrapolation.

Funding

• Not specified ( likely funded by European research programs or grants)

Citation

@article{Kemsley2026Mortality,
  author = {Kemsley, Sarah Wilson and Fromentin, Jowan L. and Pastine, Bikem and Dong, Xiaowen and Matthews, Tom and Masselot, Pierre and Meißner, Katrin J. and Perkins-Kirkpatrick, Sarah E and Slater, Louise J.},
  title = {Mortality under locally unprecedented heat is better estimated with cross-regional learning},
  journal = {Environmental Research Letters},
  year = {2026},
  doi = {10.1088/1748-9326/ae8317},
  url = {https://doi.org/10.1088/1748-9326/ae8317}
}