Current Research

Analog forecasting is a method whereby scientists predict how a current climate state will progress by analyzing how similar climate states in the past evolved. I am harnessing neural networks to help identify these similar climate states and using them to make weather/climate projections between ~2 weeks and ~3 months. 

I previously studied two types of animal motion. First, I worked on a project investigating the saltatory foraging behavior of the American Robin, before beginning my undergraduate thesis analyzing the collective motion of locust hopper bands. An example video is shown above, where color represents the locusts' motion states (stationary, crawling, or hopping).