TOOCAN is a cloud tracking algorithm, described in Fiolleau and Roca 2013, to detect and track Mesoscale Convective Systems (MCSs) from the geostationary infrared observations.
The tracking algorithm works in a volume of infrared images to identify and track MCSs not any more with the traditional detection and tracking steps but in a single 3D (spatial+time) segmentation step. That’s way the TOOCAN methodology is based on an iterative process of detection and spread of the convective seeds in order to associate the convective cores to their corresponding stratiform anvils and cirriform parts in the 3-dimensional domain. Thus, the high cold shield defined by a 235K threshold can be decomposed in several mesoscale convective systems in the spatio-temporal domain.
Here, the TOOCAN algorithm has been applied on the homogenized infrared observations from the fleet of multi-agencies meteorological geostationary satellites data (Fiolleau etal 2019) to monitor the entire tropical belt for the 2012-2016 period. Each geostationary platform permits the processing of a specific region (Eastern-Pacific, America, Africa, India, Western-Pacific).
|Platform||Nadir location||Instrument||Central wavelength||Spectral interval||Spatial resolution at nadir||Temporal resolution||Tracking region||Region name||Period|
|GOES-15||135° W||IMAGER||10.7 µm||10.2-11.2 µm||4 km||30 min||180°W-105°W 40°S-40°N||Eastern Pacific||Jan 2012-Dec 2016|
|GOES-13||75° W||IMAGER||10.7 µm||10.2-11.2 µm||4 km||30 min||111°W-30°W 40°S-40°N||America||Jan 2012-Dec 2016|
|METEOSAT-8/9/10||0°||SEVIRI||10.8 µm||9.8-11.8 µm||3 km||15 min||45°W-45°E 40°S-40°N||Africa||Jan 2012-Dec 2016|
|METEOSAT-7 (IODC)||57.5° E||MVIRI||11.5µm||10.5-12.5 µm||5 km||30 min||12°E-107°E 40°S-40°N||India||Jan 2012-Dec 2016|
|MTSAT-2||145° E||IMAGER||10.8 µm||10.3-11.3 µm||4 km||30 min||94°E-170°W 40°S-40°N||Western Pacific||Jan 2012-May 2015|
|HIMAWARI-8||140.7° E||AHI||11.2 µm||11.0-11.4 µm||2 km||10 min||94°E-170°W 40°S-40°N||Western Pacific||Jun 2015-Dec 2016|
Table I: Technical characteristics of the operational geostationary satellites fleet and the associated imagers used over the 2012-2016 period.
To get a homogenized MCS database, a homogenization procedure has been performed on the IR database (Fiolleau etal 2019):
- The temporal frequency used is 30min, even for MSG and HIMAWARI-8, to avoid over segmentation of the convective systems due to their original temporal resolution of respectively 15 and 10 minutes compared to other platforms.
- The inhomogeneity of the spatial resolution has been accounted for by regridding the IR data of each geostationary satellite from their native projections to a common map projection, a 0.04° equal-angle latitude/longitude grid.
- The IR data have also been spectrally normalized and inter-calibrated by using the ScaRaB/Megha-Tropiques observations.
- A limb adjustment has been also applied to correct the brightness temperature dependency to the viewing zenithal angle.
About 5×106 MCSs have been identified over the entire tropics for the 5-year period allowing the documentation of their morphological characteristics along their life cycles.
As a result of this processing, two databases have been built:
- a level-2 database called TOOCAN database
- a level-3 database called CACATOES database, and derived from the TOOCAN database