Preprints
https://doi.org/10.5194/egusphere-2022-179
https://doi.org/10.5194/egusphere-2022-179
 
11 Apr 2022
11 Apr 2022
Status: this preprint is open for discussion.

Intercomparison of Four Tropical Cyclones Detection Algorithms on ERA5

Stella Bourdin1, Sébastien Fromang1, William Dulac2, Julien Cattiaux2, and Fabrice Chauvin2 Stella Bourdin et al.
  • 1Laboratoire des Sciences du Climat et de l’Environnement, LSCE/IPSL, CEA-CNRS-UVSQ-Université Paris-Saclay, Gif-sur-Yvette, France
  • 2Centre National de Recherches Météorologiques, Université de Toulouse, Météo France, CNRS, Toulouse, France

Abstract. The assessment of Tropical Cyclones (TC) statistics requires the direct, objective, and automatic detection and tracking of TCs in reanalyses and model simulations. Research groups have independently developed numerous algorithms during recent decades in order to answer that need. Today, there is a large number of algorithms, often referred to as trackers, that aim to detect the positions of tropical cyclones in gridded datasets.

This paper compares four trackers with very different formulations in detail. We assess their performances by tracking tropical cyclones in the ERA5 reanalysis and by comparing the outcome to the IBTrACS observations database.

The first section of the paper finds typical detection rates of the trackers ranging from 75 to 85 %. At the same time, false alarm rates (FAR) greatly vary across the four trackers and can sometimes exceed the number of detected genuine cyclones. Based on the finding that many of these false alarms are extra-tropical cyclones, we adapt two existing filtering methods common to all trackers. Both post-treatments dramatically impact FARs, which range from 9 to 36 % in our final catalogs of tropical cyclones tracks. We then show that different traditional metrics can be very sensitive to the particular choice of the tracker, which is particularly true for the TC frequencies and their durations. By contrast, all trackers identify a robust negative bias in ERA5 tropical cyclones intensities, a result already noted in previous studies.

We conclude by advising against using as many trackers as possible and averaging the results. A more efficient approach would involve selecting one or a few trackers with well-known properties.

Stella Bourdin et al.

Status: open (until 11 Jun 2022)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse

Stella Bourdin et al.

Data sets

Primary data + analysis code Bourdin, Dulac https://doi.org/10.5281/zenodo.6424432

Stella Bourdin et al.

Viewed

Total article views: 343 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
247 89 7 343 5 3
  • HTML: 247
  • PDF: 89
  • XML: 7
  • Total: 343
  • BibTeX: 5
  • EndNote: 3
Views and downloads (calculated since 11 Apr 2022)
Cumulative views and downloads (calculated since 11 Apr 2022)

Viewed (geographical distribution)

Total article views: 329 (including HTML, PDF, and XML) Thereof 329 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 16 May 2022
Download
Short summary
Climate models output results in the form of gridded datasets. In order to study tropical cyclones, one needs objective and automatic procedures to detect their specific pattern. We study four algorithms performing this detection by applying them to a reconstruction of the climate in which we expect to find the observed storms. We conclude that these algorithms differ in their sensitivity to weak disturbances so that they provide different frequencies and durations.