Artificial intelligence can predict the weather and human health

Climate change is increasingly recognized as one of the most significant public health challenges of our time. In the United States alone, 70% of Americans faced extreme weather events in 2022. From heatwaves and droughts to hurricanes and wildfires, these events are not only growing in frequency but also in their potential to overwhelm healthcare systems.

Despite governmental and institutional efforts to mitigate climate change, adaptation to its inevitable impacts has taken a backseat. As extreme weather intensifies, the need for proactive healthcare solutions becomes ever more pressing.

Enter machine learning (ML), a tool with transformative potential for predicting health outcomes linked to climate-sensitive extreme weather. By analyzing vast datasets—including clinical records, socioeconomic factors, and environmental conditions—ML can forecast health risks at both individual and community levels.

This is not a theoretical exercise; the same predictive algorithms that guide breast cancer treatments or diagnose coronary artery disease are now being adapted to anticipate healthcare needs during climate emergencies.

However, the use of ML in predicting health impacts from climate-driven events is still in its infancy. A recent review highlighted just seven studies employing ML to forecast such outcomes. Despite this limited application, the potential for ML to revolutionize climate-related healthcare is vast.

Consider the “tripledemic” of 2023, when COVID-19, RSV, and the flu coincided with raging wildfires. Hospitals nationwide braced for an influx of respiratory illness cases. Traditional methods of planning for patient surges—relying on historical data and intuition—left hospital administrators unsure about peak times and resource allocation.

John Brownstein, Chief Innovation Officer at Boston Children’s Hospital, sought a better approach. His team integrated environmental, behavioral, and infectious disease data into a machine learning model, which provided detailed forecasts on when respiratory cases would peak. “We could predict to the day when the highest-level capacity needs would be,” says Brownstein. This level of precision allowed hospitals to manage resources efficiently, ensuring patients received timely care.

Such predictive capabilities exemplify how AI can bolster healthcare systems’ resilience against climate-related health crises. By offering granular insights, ML helps healthcare providers anticipate demand and plan accordingly.

Beyond immediate healthcare needs, AI is transforming how we understand and model climate systems. Traditional climate models rely on physics-based equations to simulate atmospheric and oceanic dynamics, but these require immense computational power and still leave gaps in long-term forecasts. AI offers a complementary approach.

For instance, Google DeepMind’s GraphCast delivers highly accurate hurricane path predictions and ten-day weather forecasts, outperforming traditional models. Similarly, Microsoft’s Aurora can forecast global air pollution patterns up to five days in advance. These tools empower healthcare providers to prepare for the health consequences of poor air quality, such as exacerbated respiratory and cardiovascular conditions.

However, challenges remain in extending AI’s predictive capabilities to long-term climate impacts. Hybrid models, which incorporate AI into conventional physics-based frameworks, are emerging as a promising solution. By grounding AI predictions in established scientific principles, researchers hope to balance the speed and efficiency of AI with the reliability of traditional methods.

One of AI’s most significant contributions to climate health is its ability to identify at-risk populations. Rising temperatures, for instance, are known to increase the likelihood of heat-related illnesses. But the relationship between climate factors and health outcomes is often more complex. AI can analyze vast amounts of data from electronic health records, insurance claims, and even social media posts to uncover patterns that might go unnoticed.

Francesca Dominici, a biostatistics professor at Harvard, underscores the importance of such tools. Her team has used AI to reveal connections between rising local temperatures and antibiotic resistance, a finding with significant implications for public health. “AI allows us to uncover relationships between multiple diseases and environmental factors simultaneously,” she explains.

This capability enables targeted interventions. For example, before a heatwave, AI models can predict which populations—such as the elderly or those with pre-existing conditions—are most at risk. Policymakers can then prioritize distributing air conditioners or opening cooling centers, potentially saving lives.

AI’s role extends beyond prediction. Initiatives like Climateverse, co-launched by emergency medicine specialist Satchit Balsari, aim to integrate and make accessible fragmented climate and health data.

In Southeast Asia, where extreme weather events frequently disrupt healthcare systems, Climateverse uses an AI-powered chatbot to help researchers and policymakers identify vulnerable communities and allocate resources effectively.

Moreover, AI could play a critical role in decarbonizing healthcare systems. By optimizing energy use within hospitals and identifying the most effective strategies to reduce carbon emissions, AI supports the dual goal of improving health outcomes and mitigating climate change. This dynamic application underscores the interconnected nature of climate and health challenges.

Despite its potential, AI’s integration into climate health raises ethical and practical concerns. AI models require significant computational power, which can increase carbon emissions. As the technology proliferates, its energy demands may offset some of the environmental benefits it seeks to provide.

Accuracy and bias are also critical issues. AI predictions must be reliable and reflect the diverse populations they serve. Dominici emphasizes the need for transparency and rigorous validation. “This balance of harnessing the good and mitigating the bad of AI is really important for us to embody at Harvard and in medicine when we’re dealing with human lives,” she says.

The path forward involves careful stewardship. As AI tools become more sophisticated, their developers must ensure they enhance, rather than undermine, climate resilience. Open dialogue and collaboration between scientists, policymakers, and healthcare providers will be essential to realizing AI’s full potential in combating climate-related health risks.

Note: Materials provided above by The Brighter Side of News. Content may be edited for style and length.

Like these kind of feel good stories? Get The Brighter Side of News' newsletter.