TOOCAN

TOOCAN is a cloud tracking algorithm, described in Fiolleau and Roca (2013), developed to detect and track Deep Convective Systems (DCSs) using geostationary infrared observations.
Unlike traditional detection–then–tracking methods, TOOCAN operates directly in a 3D (space + time) segmentation framework. The methodology is based on an iterative detection and spreading process of convective seeds, associating convective cores with their anvil clouds in the spatio-temporal domain. Within this framework, the cold cloud shield, defined by a 235 K brightness temperature threshold, can be decomposed into several distinct DCSs, tracked consistently over their entire life cycles.
Building upon this framework, the TOOCAN algorithm was applied to 10.8 μm infrared observations from METEOSAT Second Generation, at full native spatial (0.04°) and temporal (15 min) resolution, over the ORCESTRA/MAESTRO domain (30° E–60° W, 10° S–25° N) for the August–October 2024 period.
The TOOCAN–ORCESTRA/MAESTRO dataset consists of four main file types:
- Regional segmented images (NetCDF): METEOSAT native spatial resolution, 15-minute frequency.
- Regional and monthly tracking files (in ASCII and NetCDF): document integrated morphological parameters (e.g., lifetime duration, propagation distance) as well as the evolution of morphological parameters (area, min Brightness Temeprature...) through all the DCS life cycles.
- A Navigation file (navig_msg_-60_30_-10_25.nc): provides pixel-level latitude/longitude information.
- Geo-referenced shapefiles (GeoJSON): describe tropical DCSs with detailed spatial, thermal, and temporal attributes, suitable for mapping, visualization, and scientific analysis.

This TOOCAN dataset offers a consistent, high-resolution view of tropical DCS evolution, enabling comprehensive studies of convective organization and its role in tropical atmospheric processes for the ORCESTRA/MAESTRO campaign.