Mathematical Modeling of Air Pollution Over the Western US

In almost every field (and especially in any kind of data science), there will be times when we wish to estimate data we don't have by using the data that we do have. Given the plethora of machine learning algorithms and the similarly-daunting number of implementations of these algorithms (in different languages, packages, etc), it can be difficult to know where to start. This post shares my biggest takeaways from diving into using some of these tools, in the context of estimating air pollution from wildfires over the western US. Remotely sensed image of the 2015 Washington Wildfires, showing smoke mixed with cloud cover. Since September 2017 I've worked on the Environmental Health Team in Earth Lab, investigating the impacts of exposure to air pollution from wildfires on human respiratory and cardiovascular health in the western US.