I was delighted to give a talk at the Emory TMLS workshop: What have we learned from this pandemic about how (not) to model disease spread?
My talk was titled “Statistical vs mechanistic models and the climate-COVID19 link“. I discuss some of the pitfalls in applying simple statistical models to investigate the climate-COVID19 relationship and how these can be overcome. I make three broad points. First, statistical models need to have appropriate controls for trends over time and spatial biases. Second, in understanding the implications of statistical results, it is helpful to incorporate findings into a mechanistic, epidemiological model. Third, and more broadly, interdisciplinary collaboration could really improve the scope of research on this topic.
There is a youtube video of the talk here (my talk is up first): https://www.youtube.com/watch?v=vjl8ztuo9Ug&ab_channel=EmoryTMLS