Storm-centric Analysis of Tropical Cyclones in Python

Tropical cyclones, commonly referred to as hurricanes in the North Atlantic and eastern North Pacific, are scientifically fascinating storms that often harm people and property. Our knowledge of these systems has grown exponentially in the past century, yet most discoveries beget more questions. Our ability to observe tropical cyclones has also improved, particularly in the past few decades, with much of the technology driven by features we wish to better understand: more frequent observations reveal faster-acting phenomena and higher spatial resolution shows processes at smaller scales. However, the rapid increase of available data brings its own challenges with respect to storage, processing, analysis, and interpretation. For example, the Advanced Baseline Imager (ABI; operational on the GOES-East and GOES-West satellites) boasts sixteen channels ranging from 0.5- to 2-km resolution and produces a full-disk scan every 10 minutes from each channel. We can now "see" unprecedented detail of tropical cyclone structure and evolution with these images, yet these large and numerous files are not necessarily easy to use, let alone analyze.

Open source software developers, many of them with meteorology expertise, have provided invaluable Python tools that enable us to learn more about atmospheric phenomena such as tropical cyclones. By using only Python modules, it is now possible to remotely access big datasets -- even those that require login credentials -- to extract storm-centered grids for only the region of interest and then diagnose common patterns within those fields. Such tools have revolutionized my research and made new discoveries possible.

In this presentation, I will share progress on how I use Python modules including Siphon, MetPy, cartopy, and xarray to explore tropical cyclone features revealed by ABI data and assess the influence of environmental conditions on those features. All tropical cyclones are evaluated within a storm-centric framework to best quantify the environment's influence on these systems. Fields such as wind, temperature, humidity, and vertical velocity are provided by the European Centre for Medium-Range Weather Forecasts' 5th generation reanalysis (ERA5) and hosted by the University Corporation for Atmospheric Research (UCAR). Since ABI coverage begins in 2017, cases are limited to the 2017-2019 hurricane seasons, though results that include earlier years using coarser datasets will be shared if presentation time allows. This work should not only produce additional insights into tropical cyclone behavior but also develop visualizations that can be used in classrooms and to share the value of our research with those outside the scientific community.