Background/Objectives

While we know that climate change will impact the risk of infectious diseases for people, animals, and crops, and we understand some of these impacts from a broad-stroke statistical perspective, we still need tools that provide robust predictions and what-if scenarios across continents at a fine enough scale to impact local decision making. This requires mechanistic models that can incorporate ecosystems, human populations/infrastructure, and relevant animal or vector populations along with pathogen transmission models. We developed a multi-scale and generalizable framework to couple global earth-systems models with disease transission and population models to enable fine-scale quantifying of future risk that includes mitigations and scenario testing. Such models, and the required calibration/validation, are directly dependent on multi-disciplinary teams working closely together.

Approach/Activities

We applied our framework to two test cases: West Nile virus and dengue. Both are mosquito-borne pathogens that are impacted directly by climate and habitat/ecology. West Nile virus spreads primarily among birds and mosquitoes, requiring a wildlife interface, while dengue is primarily spread between humans and mosquitoes. Both test cases require accurate mosquto population dynamics models as driven by multiple ecosystem and weather/climate drivers as well as land use and human populations. To drive the models into the future, we used DOE's E3SM Earth System model and developed a tiling spatial method (Eco-Pop Units) to enable prediction at the appropriate scale for the mosquito-borne disease systems we were considering. This method is driven by a machine-learning algorithm that groups "like" mosquito habitat and human interfaces to enable transfering information from data-rich to data-poor regions.

Results/Lessons Learned

We demonstrated that our model can reproduce past dynamics (from mosquito trap data and disease case counts) accurately, capturing the non-linear interactions between factors including temperature, hydrology, precipitation, bird migration, etc. This is the foundation to the predict into the future under different climate scenarios, to enable short-term predictions given weather forecasting, and to enable exploration of the impacts of mitigation strategies. In the future we will apply this framework to other case studies, including crop pathogens and tick-borne pathogens.