An Algorithm for the Detection and Tracking of Tropical Mesoscale Convective Systems Using Infrared Images From Geostationary Satellite

This paper focuses on the tracking of mesoscale convective systems (MCS) from geostationary satellite infrared data in the tropical regions. In the past, several automatic tracking algorithms have been elaborated to tackle this problem. However, these techniques suffer from limitations in describing convection at the "true" scale and in depicting coherent MCS life cycles (split and merge artifacts). To overcome these issues, a new algorithm called Tracking Of Organized Convection Algorithm through a 3-D segmentatioN has been developed and is presented in this paper. This method operates in a time sequence of infrared images to identify and track MCS and is based on an iterative process of 3-D segmentation of the volume of infrared images. The objective of the new tracking algorithm is to associate the convective core of an MCS to its anvil cloud in the spatiotemporal domain. The technique is applied on various case studies over West Africa, Bay of Bengal, and South America. The efficiency of the new algorithm is established from an analysis of the case studies and via a statistical analysis showing that the cold cloud shield defined by a 235-K threshold in the spatiotemporal domain is decomposed into realistic MCSs. In comparison with an overlap-based tracking algorithm, the analysis reveals that MCSs are detected earlier in life cycle and later in their dissipation stages. Moreover, MCSs identified are not anymore affected by split and merge events along their life cycles, allowing a better characterization of their morphological parameters along their life cycles.