Climate model representation of January 2025 conditions using nextGEMS. Watch the simulation here.

 

By Cosimo Enrico Carniel
Master’s Student
ETH Zürich Institute for Atmospheric and Climate Science

Extreme heat events, characterized by prolonged periods of unusually high temperatures, are emerging as one of the most pressing consequences of climate change. These events, which pose significant threats to human health, ecosystems and infrastructure, including military ones, have garnered increasing attention, particularly as their frequency and intensity escalate. Such extreme heat events are particularly dangerous in regions where populations are highly vulnerable to heat stress and the associated health risks, such as India, Pakistan and the Southern Mediterranean Basin. In regions where military tensions are at stake, such phenomena affect the functionality of equipment on land, reducing aircraft performance, limiting payload capacities, and disrupting naval systems by increasing maintenance needs or lowering operational efficiency. No area is truly safe when extreme events occur, as shown by the 2003 European heat wave that tragically claimed more than 70,000 lives, making it one of the deadliest in recent history.

Over the past few decades, research has highlighted a significant upward trend in the occurrence of extreme heat across various regions. For instance, in the Mediterranean region, under an intermediate emissions scenario, heat waves are projected to extend 27 to 67 days per summer by the end of the century, a dramatic increase compared to the mid-21st century, when this value was estimated at six to 24 days. Such projections are alarming, especially considering the severe health risks associated with them. Indeed, human tolerance to heat, already challenged by global warming, is further compromised when both temperature and humidity rise simultaneously. This is particularly concerning in regions where access to air conditioning and other cooling measures is limited, and where public health infrastructures may not be equipped to handle large-scale heat emergencies.

In addition to direct impacts on human health, extreme heat events also exacerbate environmental hazards, such as wildfires. Heat waves can lower the moisture content in plants and soils. Coupled with drought conditions, this creates ideal conditions for wildfires to start and spread rapidly. These fires release massive amounts of carbon dioxide into the atmosphere, further contributing to climate change and amplifying the feedback loop of warming temperatures and more frequent extreme heat events.

One of the key innovations in the study of extreme heat events is the use of wet bulb temperature (TW) or wet bulb globe temperature (WBGT) as a useful indicator for alert thresholds. TW combines temperature and humidity, while WBGT takes into account radiation and wind speed to provide a more accurate measure of human heat stress. Many studies highlight 35° C TW as the threshold at which human survival becomes impossible due to the failure of natural cooling mechanisms such as sweating. However, research increasingly shows that true risk to life begins at much lower TW values. In certain contexts, for example, in the military, strenuous activity is restricted when TW reaches around 31° C, and soldiers need to follow strict hydration and rest protocols. This makes the assessment of TW (or WBGT) vital for evaluating the impact of extreme heat events, particularly in regions with high humidity.

Next-Generation Models and the Role of SRMs

Traditional climate models have been instrumental in identifying broad trends in global and regional climate patterns, but they are limited when applied to localized extreme weather events. Advances in climate modeling now offer new opportunities to quantify changes in extreme heat events. At the forefront of these advancements are the storm-resolving models (SRMs), which operate at high spatial resolutions (between 2 and 10 km). These models capture small-scale atmospheric and oceanic processes critical for understanding extreme weather, including local manifestation of heat events. SRMs, coupled with high-resolution oceanic models, represent a significant leap forward from traditional climate models, which often miss localized coastal and atmospheric phenomena that contribute to extreme heat dynamics. SRMs can potentially better capture complex interactions such as sea breeze circulation, convective processes, land-sea temperature contrasts, mesoscale ocean eddies, and sea ice dynamics, which are critical for understanding heat waves, particularly in coastal and topographically complex regions.

For instance, in coastal areas, the interaction between land and sea can significantly influence temperature extremes, with processes such as coastal upwelling and sea breeze circulation playing crucial roles. SRMs are better equipped to capture these dynamics, making them invaluable tools for studying extreme heat in these regions.

To address these challenges the Next Generation Earth Modelling Systems (nextGEMS), a European Union Horizon 2020 initiative, has developed two storm-resolving, fully coupled Earth system models. The goal of these models is to provide climate services through the Climate Change Adaptation Digital Twin developed in the EU’s Destination Earth initiative (DestinE). By leveraging these models, the nextGEMS project aims to improve our understanding of the physical processes driving extreme heat events while addressing persistent biases in conventional Earth system models.

As of March 2024, nextGEMS has produced 30-year climate projection simulations at kilometer scales. A Hackathon, where the produced simulations are carefully examined by several groups of scientists, is scheduled for March 2025.

The evaluation of these models through the integration of observational data and reanalysis products is now the critical step. Reanalysis data, such as provided by the ERA5 data set, offer a comprehensive historical record of atmospheric and surface conditions and serve as a reference for evaluating model performance in simulating extreme heat. Observational data from weather stations also provide essential validation of model outputs. The combination of model simulations, reanalysis data, and observational records enables researchers to assess how well models capture the frequency, intensity, and spatial distribution of extreme heat events.

The Future of Extreme Heat Modeling

While significant progress has been made in developing next-generation climate models, much work remains to predict extreme heat events at the needed spatial and temporal resolutions. Further research is necessary to improve the representation of key physical processes in these models, especially concerning land-sea interactions, convective processes and regional topography—all of which are crucial for understanding heat wave development.

Enhancing the resolution of models, particularly for long-term projections, is essential to accurately predict future occurrences of extreme heat under different climate scenarios. In this context, artificial intelligence (AI) offers a powerful approach to advance our understanding of extreme heat events. AI tools can analyze vast amounts of data quickly and identify patterns that may not be immediately apparent through traditional methods. Machine learning algorithms, for instance, can be trained to recognize precursors to extreme heat events by integrating observational data, reanalysis products and model outputs. AI-driven approaches can also improve the accuracy of predictions in regions with complex terrain or sparse observational data, and assist in directly simulating climate processes at the local level. This enhanced precision will offer policymakers and planners more actionable information to mitigate the impacts of extreme heat.

By improving our ability to predict extreme heat events, we can develop better tools to prepare for the impacts of climate change, particularly in regions with an urgent need for adaptation strategies, and in support of timely and effective military operations.

Learn more here.