Project Scientist I, National Center for Atmospheric Research



TempestExtremes exampleTempestExtremes is a suite of flexible detection and characterization algorithms developed for processing large climate datasets. This package uses an algorithmic framework known as "MapReduce" to first detect candidate events at individual times using specified criteria. Stitching is then used to assess the evolution of related detections over time. The result is an objective calculation of the climate indicator that can be automated and parallelized for multiple datasets. Generalized kernels (such as defining local minima/maxima, contours, radial integrals, areal/closed contours, etc.) are available as "building blocks," allowing users to easily define algorithms on the command line. Structured and unstructured NetCDF meshes are supported natively. TempestExtremes has been used for applications such as quantifying tropical cyclones, extratropical cyclones, snowstorms, mesoscale convection, atmospheric rivers, and blocking events in climate data and verifying hurricane forecasts.

TempestExtremes (and example scripts/data) is available on Github.


ExtraTrack exampleThe Extratropical Transition (tropical cyclone) Tracker (ExTraTrack) is a wrapper codebase that calculates Cyclone Phase Space (CPS) parameters from tropical cyclone trajectories and associated gridded data. Given pointwise trajectories (from either an observational product such as IBTrACS or algorithmic software, such as TempestExtremes), the code extracts thermal symmetry and shallow/deep warm/cold core metrics from gridded netCDF data. The software follows storms following the termination of a tropical trajectory, allowing the full lifecycle of a tropical cyclone to be assessed, from tropical cyclogenesis, to onset of extratropical transition and eventual evolution to a fully cold-core, baroclinic system. Structured, CF-compliant, NetCDF data are supported, but functions used to calculate CPS values are also provided as standalone NCL, allowing for users to modify driver code to suit different data needs if necessary.

ExTraTrack (and example scripts/data) is available on Github.