Extreme heat is a leading weather-related threat to human health, with risk projected to intensify as the climate warms. Robust and transferable estimates of heat-related health risk require an understanding of how socioeconomic and demographic contexts influence who is most vulnerable in addition to the role confounding environmental stressors such as humidity and air pollution play in exacerbating dangerous conditions. This session brings together environmental epidemiologists, data scientists, climate scientists, and human physiologists to advance methods for predicting and forecasting heat-related morbidity and mortality, and to better understand the factors that might modify risk beyond temperature.

We seek machine-learning, statistical, physiological or hybrid approaches that provide new insights into heat-related health risks and/or translate these insights into improved predictive frameworks. Emphasis is placed on novel methodologies improving the validation and transferability of models across regions, climates, and populations. We are particularly interested in approaches that address limited or heterogeneous health data to enable prediction in data-sparse settings, where populations highly vulnerable to extreme heat may be poorly represented. These include physiologically based approaches that derive limits of human heat tolerance from thermoregulation rather than from local health records, and that resolve risk across distinct population cohorts. We also welcome contributions improving predictions under anthropogenic warming, extending model transferability into different climates. Advances in our understanding of compound risks, prediction methods and model transferability can support more effective and equitable Heat-Health Warning Systems.