Identifying vulnerable populations and transmission pathways by geographic correlation of the environment to human health

The social determinants of individuals' health (e.g., socio-economic, demographic, and genetic conditions) play a major role in the health of an entire population. However, in comparison to environmental data, global data on the social determinants of health is spatially coarse, infrequently updated, and costly to measure. From global mapping efforts of the recent COVID-19 pandemic it is clear that social data is not meeting the fine spatial quality needed for mapping vulnerable populations and transmission pathways. Most maps produced generalized to larger administrative units (such as counties, states), and have not identified distinct areas of vulnerable populations apart from the surrounding environment where no population resides. We present a framework that uses environmental determinants of health, instead of social ones. Other studies that link the environment to human health have done so by analyzing one ecosystem service (such as clean air) to the health of the population. Instead of relating one ecosystem service to the health of the population, this framework breaks the environmental features that produce the ecosystem service into parts (forest, temperature, precipitation). Each feature is then related to human health. With the amount of data available it is feasible to include change in monitored features over time, and create predictors for the impact of the change of monitored features on the health of populations. This framework generalizes ecosystem services and disservices into one value that an environmental feature provides. This helps to manage uncertainty of how an individual ecosystem service affects health. Application of this framework will allow for fine scale monitoring of vulnerable populations and transmission pathways of various infectious diseases. This framework is particularly relevant to newly emerging infectious diseases, such as COVID19, whose socially determinant risk factors are unknown (or data scarce) and to which we have to respond in a rapid manner.