Interdisciplinary approach to include the effect of weather and climate in models for the spread of airborne infectious diseases

Question. Interdisciplinary research is becoming an increasingly important aspect of climate and weather impact research. In this work, we present a case study to include high-resolution climatology in models for the spread of airborne infectious disease. 

Approach. To achieve this, we expand the collaboration between two institutions, the National Institute for the Public Health and the Environment (RIVM) and the Royal Netherlands Meteorological Institute (KNMI). In this research, we bring two developments together. The RIVM is focusing on modelling the spread of airborne infectious diseases and assess their risk to human health, while the KNMI is focusing on providing high-resolution weather and climate maps by blending official KNMI weather station data, crowd-sourced Weather Observation Website (WOW, https://wow.knmi.nl) data and static information like land use and population density. 

On a technical level, the high-resolution weather maps are based on multi-fidelity Bayesian regression kriging. The core of the disease spread model is a spatial transmission kernel, which estimates the probability of infection as a function of the distance from the source. However, in this basic form, this approach has some limitations, because it does not take into account the impact of other important factors as for instance meteorological and environmental variables. In the combined effort by RIVM and KNMI we aim to improve and extend the current framework of the disease spread model by including the high resolution weather maps into the spatial transmission kernel.  Importantly, since both the weather maps and kernel model quantify uncertainty we can properly propagate uncertainty by Bayesian principles in the disease model. Our hypothesis is that, when we bring these two lines of research together, the high-resolution weather maps can lead to better models of disease spread, by including crucial covariates. In addition, such research is an important opportunity to establish links between science and society, both on the crowd-sourced weather observations side as on the publicly relevant side of health impact. 

Preliminary results. Currently, we are developing the methodology to improve and combine these two approaches from different disciplines. In this presentation, we will present this methodology for combination of approaches, as well as preliminary results from the two lines of approach that we aim to combine.