Climateverse: Making Climate Data Actionable
Satchit Balsari, Associate Professor of Emergency Medicine (Harvard Medical School)
Caroline Buckee, Professor of Epidemiology (Harvard T.H. Chan School of Public Health)
This project will implement a pilot project in Kerala, India, to integrate climate data into disaster response and environmental decision-making, utilizing a tool called Climateverse. This project involves scaling up Climateverse, customizing a chatbot for non-experts, partnering with local agencies to develop best practices, aiming to apply these strategies in other regions of the global south.
Data Platforms, Causal inference (CI) and Machine Learning Methods (ML) for Identifying Social Determinants of Health in the US
Francesca Dominici, Clarence James Gamble Professor of Biostatistics, Population, and Data Science (Harvard T.H. Chan School of Public Health)
This project will develop data architectures, enhance casual inference algorithms and software, and develop interpretable machine learning mechanisms to advance evidence-based policies and actions to protect populations under several converging trends: the co-occurrence of multiple exposures to extreme weather and air pollution, the rapidly aging US population, and societal stressors such as gun violence.
Dramatically Accelerating forward Bayesian Modeling for
Petabyte Data Sets through AI Compiler Technologies
Peter Galison, Joseph Pellegrino University Professor (Harvard University)
AI-driven compiler technologies will transform scientific imaging software to achieve major speedups in black hole imaging pipelines required for next-generation observatories like BHEX. Deliverables include optimized open-source software and a high-performance imaging pipeline that can be reused across astronomy and other data-intensive scientific domains.
Linking climate and environmental exposures to malnutrition in a food- insecurity hotspot
Christopher Golden, Assistant Professor of Nutrition and Planetary Health (Harvard T.H. Chan School of Public Health)
In this project, data affecting multiple sectors in Madagascar will be triangulated and harmonized so that it can be accessed in real-time by multiple government ministries to analyze and respond to health, environmental, and agricultural challenges.
Developing Climate-Smart Public Health Systems in Nepal
Christopher Golden, Assistant Professor of Nutrition and Planetary Health (Harvard T.H. Chan School of Public Health)
A national climate–health data platform will integrate environmental, satellite, and health-system data to quantify how climate change drives infectious disease and malnutrition across Nepal. The work produces scalable data pipelines and analytical tools to support government decision-making, early warning systems, and climate-resilient public health policies.
Climate change and volatility in food supply: historical contributions to and future implications for food insecurity
Peter Huybers, Professor of Earth and Planetary Sciences, Environmental Science and Engineering (Harvard Faculty of Arts & Sciences)
This project will study how climate change increases volatility in food production and the implications of variability in supply for food security, addressing existing limitations in data and methodologies.
Global quantification of methane emissions at 25-km resolution by inversion of TROPOMI satellite observations using massively parallel cloud computing
Daniel Jacob, Vasco McCoy Family Professor of Atmospheric Chemistry and Environmental Engineering (Harvard Faculty of Arts and Sciences)
A massively parallel AWS-based inversion framework will combine TROPOMI satellite observations with a high-performance atmospheric transport model to estimate global methane emissions at 25-km resolution. The resulting global emissions dataset and Jacobian data cube enable researchers and policymakers to assess emission sources, conduct regional inversions, and guide mitigation strategies.
Development of AI predictive tools from data-based
simulations to unravel how epigenetic mechanisms control human health and disease
Efthimios Kaxiras, John Hasbrouck Van Vleck Professor of Pure and Applied Physics (Faculty of Arts and Sciences)
A multiscale simulation and AI framework will link quantum chemistry, molecular dynamics, and machine-learning models to reveal how complex histone post-translational modifications regulate liquid–liquid phase separation in chromatin. The effort generates large simulation datasets, an AI predictive model, and a structured data cube to accelerate biomedical discovery and therapeutic target identification.
Physics-aware foundation models for extreme atmospheric events
Petros Koumoutsakos, Herbert S. Winokur, Jr. Professor of Computing in Science and Engineering (Harvard John A Paulson School of Engineering and Applied Sciences)
Frank Keutsch, Stonington Professor of Engineering and Atmospheric (Harvard John A Paulson School of Engineering and Applied Sciences)
This project will develop a physics-aware foundation model that leverages observational datasets of atmospheric conditions to enhance existing physical models in order to drastically improve forecasting abilities for extreme atmospheric events.
Development of a deep learning model to enable serosurveillance of vector-borne diseases and their hosts
Daniel Neafsey, Associate Professor of Immunology and Infectious Diseases (Harvard T.H. Chan School of Public Health)
Sarah Fortune, John LaPorte Given Professor of Immunology and Infectious Diseases (Harvard T.H. Chan School of Public Health)
Junwei Lu, Assistant Professor of Biostatistics (Harvard T.H. Chan School of Public Health)
The immune system generates antibodies in response to infection by vector-borne pathogens. This project will link the presence and relative levels of these antibodies to climate change metrics to elucidate connections between climate change and vector-borne diseases. The team will then create a surveillance tool that can generate large-scale prevalence data to inform public health measures, train ecological models of climate change, determine new vaccine targets for vector-borne diseases, and more.
Multi-Modal AI for Highly Personalized and Dynamic Causal Decision-Making in Cancer Treatments
Soroush Saghafian, Associate Professor of Public Policy (Harvard John F Kennedy School of Government)
Developing a multimodal AI system—built on large language models and causal reinforcement learning—to generate highly personalized, adaptive treatment recommendations for stage III melanoma patients using de-identified clinical, imaging, genomic, and social determinants of health data. Working with Dana-Farber Cancer Institute, the team will create algorithms, software pipelines, and fine-tuned LLMs capable of both prescribing optimal treatment sequences and providing physician-level explanatory reasoning.
Optimizing Global Nutrition from Seafood Harvested from a Warming Ocean
Elsie Sunderland, Fred Kavli Professor of Environmental Chemistry (Harvard Faculty of Arts and Sciences)
Christopher Golden, Assistant Professor of Nutrition and Planetary Health (Harvard T.H. Chan School of Public Health)
This project will leverage several unique, global databases to develop regional scenarios for nutritionally optimized seafood harvests that minimize toxicants and maximize micronutrients in seafood in a warming ocean.
Creating a Repository of Gene Regulatory Networks and a “Serendipity Engine” using TOPMed to accelerate Social Determinants of Health Research
John Quackenbush, Professor of Computational Biology and Bioinformatics, (Harvard T.H. Chan School of Public Health)
The team is turning NIH TOPMed’s genomic and health data into a free resource that maps gene-regulation networks for individuals and links them to clinical records and social factors that influence health. They will pre-compute all-vs-all correlations across these variables to power an online “serendipity engine” that helps researchers spot surprising patterns worth testing.
Food Power and Food Insecurity
David Yang, Professor of Economics (Harvard Faculty of Arts and Sciences)
One countries’ food power – the ability to strategically restrict others’ access to food or flood the market to reduce others’ revenue – can simultaneously support domestic food security and international food insecurity. This project will quantify the causes and consequences of food power. Which countries have food power over which other countries? Does food stockpiling lead to the accumulation of food power? And how does food power shape global patterns of food insecurity and geopolitical alignment?