WCSI - Warm Core, Symmetric, and Intensifying

A set of criteria for identifying tropical cyclones from a set of cyclone tracks

This repository contains a pip installable command line script and python function for applying the WCSI criteria to a set of cyclone tracks. Also included are a set of scripts and functions to reproduce the analysis and figures from the paper describing the method (to be submitted).

Install

git clone https://github.com/Huracan-project/wcsi.git
cd wcsi
pip install .

Applying WCSI criteria

Command line script

To apply the WCSI criteria to your tracks, you can run the command line script

wcsi filename_in filename_out

where filename_in is the name of a file that can be loaded with huracanpy.load

This will create two files

• filename_out.nc - The subset of tracks that are WCSI, with an is_tc variable added that marks the WCSI points
• filename_out.parquet - A summary table of all the input tracks showing which track IDs were identified as WCSI or not

Various different filters and thresholds can be applied. Run

wcsi --help

To see the full set of options

Note that the script assumes you have certain named variables in your file

• The cyclone phase space parameters named as cps_b, cps_vtl, cps_vtu
• relative_vorticity with pressure as a vertical coordinate

If you want to rename variables without creating a new file, you can use the python function instead

Python function

To apply the WCSI criteria to tracks using the python function you can follow the example below

import huracanpy
from wcsi import nature

# Load tracks with huracanpy
tracks = huracanpy.load("filename_in")

# Apply WCSI criteria
wcsi_tracks, summary = nature.wcsi(tracks)

# Save the tracks and summary table
huracanpy.save(wcsi_tracks, "filename_out.nc")
summary.to_parquet("filename_out.parquet")

As with the command line script, various different filters and thresholds can be applied. Run

from wcsi import nature

help(nature.wcsi)

to see the function documentation.

Paper

This repository contains all the code needed to reproduce the results in the paper. To do that, first download the ERA5 and JRA3Q tracks, and the summary table from zenodo. To produce these files I have already run

# Apply WCSI criteria to ERA5 and JRA3Q tracks
wcsi ERA5_all.nc ERA5_WCS --intensification_threshold=None
wcsi ERA5_WCS.nc ERA5_WCSI
wcsi JRA3Q_nolat-tcident.nc JRA3Q_nolat-tcident_WCSI

# Create table summary tracks by filter and matching across datasets
python -m wcsi.scripts.add_nature
# Add a nature tag to the ERA5 tracks
python -m wcsi.scripts.combine_summaries

Then, to recreate the figures in the paper

1. Download and subset IBTrACS (wcsi.scripts.ibtracs_subset)
2. Create a file with MSLP at maximum intensity matched across different datasets (wcsi.scripts.matched_lmi)
3. Run the scripts in wcsi.plot