Releases: alexcrawford0927/cyclonetracking
v13.2
The main updates from version 12.4 to 13.2 involve the outputs up the main data frame in the cyclonetrack objects. Namely,
- Changed u and v variables to be velocity instead of speed.
- Removed Dx and Dy.
- Changed sorting of main data frame to no longer always be alphabetical.
- Added the SLP gradient measure (p_grad) used to limit cyclone detection as a variable.
- Replaced all "longs" with "lons" when referring to longitude, making it the same number of characters as “lats” for latitude.
- The dist2lon function was changed so that if the input is negative, the output will be negative. (It previously gave an absolute value).
- Added a try/except statement to cTrack2sTrack function so that it will ignore cases where tids between two months don’t align instead of breaking. Such cases lead to there being two separate system tracks despite a regenesis event being detected. This is a fringe case that probably effects fewer than one in 100,000 tracks.
v12.4.1
This release is mostly to respond to a change in how the reprojection functions used have changed. Updates were also made to how tracks results are aggregated into gridded fields. None of these updates change the number of tracks generated in the test data (557 tracks in Jan 1979 of ERA5 data at 100 km resolution and default settings -- 166 of which last more than 24 hours).
Full details
-
Reprojection functions used no longer extrapolate beyond range of input grid. Therefore, reprojection code has been updated to prevent seams developing when going from rectilinear projections to polar projections. For polar grids, cells in the opposite hemisphere are now NaN instead of being unrealistically extrapolated.
-
The aggregation scripts have been updated so that if a prior aggregation script already exists (e.g., 1979-2020), an extra year of data (e.g., 2021) will be appended to the pre-existing netcdf instead of re-processing all years.
-
Climatology calculations of aggregated fields have been moved to a separate script.
-
A regional mask I often use is now available as a netCDF file on the associated Google Drive.
-
A bug in the 360-day calendar in the timeAdd function has been fixed.
Version 12.4
This release is associated with the following peer-reviewed article:
Crawford, A. D., Schreiber, E. A. P., Sommer, N., Serreze, M. C., Stroeve, J. C., & Barber, D. G. (2021). Sensitivity of Northern Hemisphere Cyclone Detection and Tracking Results to Fine Spatial and Temporal Resolution using ERA5. Monthly Weather Review. https://doi.org/10.1175/mwr-d-20-0417.1
Version 12.2
This is the pre-release version of the algorithm associated with a manuscript that is not yet published. Further modifications are possible prior to publication.