Menu Close

TOOCAN database

The TOOCAN database version 2.08 is a level-2 product available over the entire tropics with a latitudinal extension from 40° S to 40° N over the 2012-2020 period. The dataset has been updated in December 2023.
It gives an access to the integrated morphological parameters of each Deep Convective Systems (Location and time of initiation and dissipation, lifetime duration, propagated distance, maximum extent…). The DCSs are also described by their morphological characteristics at each 30min-step of their life cycles (Cold cloud surface at different level of brightness temperature thresholds, eccentricity, instantaneous velocity… )

The TOOCAN database version 2.08 is available over the 9-year period for each region of interest (Eastern-Pacific, America, Africa, India, Western-Pacific) in the following link: https://doi.org/10.14768/1be7fd53-8b81-416e-90d5-002b36b30cf8

The TOOCAN database is composed by two types of files:

  • Regional segmented images at a 0.04° spatial resolution and a 30 minute temporal frequency (in NETCDF).
  • Regional and monthly tracking files (in ASCII and in NETCDF) documenting the DCS integrated morphological parameters and the MCS parameters at each 30minute-step of their life cycles.

Each DCS is then described by a unique label, which is useful for making the link between a given DCS described in the tracking files and the same DCS identified in the segmented images. Thus, using both the tracking file and the segmented images provides access to all the pixels of a given DCS identified in the segmented images, as well as to its morphological parameters in the monthly tracking files.


The TOOCAN database version 2.08 differs from the previous version 2.06 (DOI https://doi.org/10.14768/20191112001.1) in several ways:
    – The TOOCAN algorithm identifies convective systems by working in a spatio-temporal volume of infrared images. The first brightness temperature threshold is still set at 190K to detect the convective seeds in the volume of IR images, and the the iterative detection and dilation process is stopped when the 235K threshold is reached, corresponding to the high cold cloud shield boundaries. However, to improve the quality of convective system segmentation, the detection step has been reduced from 5K to 2K in the iterative detection and dilation process.
    – New classifications of convective systems have been established, according to the definition of MCC, PECS, MβCCS, and MβECS defined in Jirak etal (2003) and Maddox (1981).
    – Also, taking advantage of the IBTrACS database (Knapp etal 2010), convective systems located within a radius of 1000km from a cyclone have been flagged.
    – The monthly tracking files have been saved in a NETCDF format in addition to the traditional ASCII format.
    – The TOOCAN segmented images in a NETCDF format have been supplemented with the information on the scanning time of the geostationary satellites.

Animation of the DCS identified and tracked by TOOCAN from the segmentation of the HIMAWARI-8 IR data over the Western Pacific region in October 2015. (Each color corresponds to a unique MCS, and the black lines represent the displacements of the MCS barycenters)

DCS Integrated morphological parameters

DCS morphological parameters along their life cycles

The technical aspects of the TOOCAN database can be found in the following documentation:

CC BY 4.0 : Permission is granted to use these data in research and publications mentionning the dataset DOI https://doi.org/10.14768/20191112001.1 and the reference articles and accompanied by the following statement: The authors acknowledge the data center ESPRI/IPSL for providing access to the data.