Cooperative Institute for Research in the Atmosphere

Quantifying tropical cyclone behavior with open-source tools across spatial and temporal scales

The ever-expanding volume of readily-available atmospheric data requires matching improvements in computational power and computing tools. There is a growing ecosystem of geoscience Python packages and increasing federal funding for such efforts, but it can be challenging to identify the most appropriate tools for a given task and then navigate the learning curve required to implement a chosen tool. Recent collaborative work has investigated seasonal tropical cyclone (TC) activity—such as the record-breaking 2020 Atlantic hurricane season—and machine-learning-friendly metrics to quantify TC convective structure with potential for short-term structural prediction. This talk will present research findings and discuss how the research was conducted by providing workflow examples and linking to sample code (http://arashi.geosci.msstate.edu/python/PythonResources.html) to support others in implementing similar approaches. It will also describe these tools’ applicability to real-time analysis by presenting data visualizations generated for the 2022 North Atlantic and eastern North Pacific hurricane seasons.