AI-driven Early Warnings System for Compound Environmental Health Risks in Europe
In recent years, rapid advances in artificial intelligence have revolutionised atmospheric forecasting, giving rise to AI-based weather prediction models (AIWPs) that collapse forecast runtimes from hours on supercomputers to minutes on a single GPU, while rivalling the skill of state-of-the-art numerical weather prediction models (NWPs). Yet this disruptive capacity remains largely untapped in public health, even as Europe faces escalating extremes of heat and air pollution. Crucially, hazards no longer occur in isolation: compound heat-air-pollution events are rising in frequency, duration and severity, disproportionately amplifying health inequalities. However, current health early warning systems (HEWS) still treat temperature and air pollution separately and issue uniform alerts, leaving those most at risk under-informed. It remains completely unexplored whether HEWS can deliver equally accurate warnings across populations under compound exposures. AI-4-HEALTH addresses this gap by demonstrating the potential of AIWPs to power operational HEWS for compound hazards, enabling faster government and public responses at unprecedented efficiency gains and a minimised carbon footprint, a breakthrough unattainable with conventional NWPs. AI-4-HEALTH will integrate AIWPs with epidemiological models to issue subgroup-specific health alerts, and for the first time, assess the forecast skills by sociodemographic strata, revealing inequalities in warning accuracy between population groups. This interdisciplinary effort directly addresses the urgent need for robust health responses to climate threats. It is strategically aligned with the Horizon Europe Work Programme, in particular Cluster 5 (Climate, Energy and Mobility) and the EU Mission “Adaptation to Climate Change.” Successful delivery will strengthen Europe’s capacity to confront climate and air-pollution challenges in a way that promotes sustainability, equality, diversity, and inclusion.
4/6 complete
Imported from OpenAIRE Spain | Grant 101277999.
Open sourceOverview
Grant agreement 101277999 | AI-4-HEALTH
Open sourceIn recent years, rapid advances in artificial intelligence have revolutionised atmospheric forecasting, giving rise to AI-based weather prediction models (AIWPs) that collapse forecast runtimes from hours on supercomputers to minutes on a single GPU, while rivalling the skill of state-of-the-art numerical weather prediction models (NWPs). Yet this disruptive capacity remains largely untapped in public health, even as Europe faces escalating extremes of heat and air pollution. Crucially, hazards no longer occur in isolation: compound heat-air-pollution events are rising in frequency, duration and severity, disproportionately amplifying health inequalities. However, current health early warning systems (HEWS) still treat temperature and air pollution separately and issue uniform alerts, leaving those most at risk under-informed. It remains completely unexplored whether HEWS can deliver equally accurate warnings across populations under compound exposures. AI-4-HEALTH addresses this gap by demonstrating the potential of AIWPs to power operational HEWS for compound hazards, enabling faster government and public responses at unprecedented efficiency gains and a minimised carbon footprint, a breakthrough unattainable with conventional NWPs. AI-4-HEALTH will integrate AIWPs with epidemiological models to issue subgroup-specific health alerts, and for the first time, assess the forecast skills by sociodemographic strata, revealing inequalities in warning accuracy between population groups. This interdisciplinary effort directly addresses the urgent need for robust health responses to climate threats. It is strategically aligned with the Horizon Europe Work Programme, in particular Cluster 5 (Climate, Energy and Mobility) and the EU Mission “Adaptation to Climate Change.” Successful delivery will strengthen Europe’s capacity to confront climate and air-pollution challenges in a way that promotes sustainability, equality, diversity, and inclusion.