Spatio-temporal characteristics of mesoscale convective systems over northwestern South America using a new persistence tracking technique.

Abstract

ABSTRACT : Mesoscale Convective Systems (MCSs) are the largest form of deep convection and play a significant role in the large-scale tropical circulation and hydrological cycle. MCSs represent less than 1% of the total number of organized convection but contribute to a significant fraction of tropical rainfall (more than 50% of total precipitation) and pose threats to society by generating severe weather events. Northwestern South America (NwSA) is characterized by a complex terrain with high MCSs activity and has some of the rainiest places on Earth. However, due to the lack of continuous information in time and space, MCSs characterization represents a challenge. Several studies performed characterizations of MCSs in this region, mainly following a Eulerian approach, focusing on the spatial characteristics but limited in the temporal features such as movement and evolution. Here we build a persistence algorithm to track MCS in a lagrangian approach to study climatological (20-yr period) characteristics of MCSs over NwSA. The Algorithm for Tracking Convective Systems (ATRACKCS) uses two high-resolution satellite datasets to identify and track MCSs: brightness temperature data (Tb) and precipitation data (P). ATRACKCS selects MCSs based on (P) and (Tb) thresholds along with extent and duration criteria and tracks them over time using an area overlapping method. We focus on the spatio-temporal variability of three main MCSs characteristics: occurrence, life cycle and propagation. Our results show clear contrast between land and ocean characteristics, like smaller MCSs with shorter life cycles that initiates in afternoon hours over land, and big long lasting and nocturnal MCSs over the oceans. We identified regions with a higher density of MCSs tracks, such as the Colombian Magdalena valley, the Colombian Pacific coast and Venezuelan Guianas Highlands. These, here called ‘hotspots’, have substantial differences with the adjacent continental region. We analyze the differences in duration, area, movement, seasonality and diurnal cycle across the study area, and associated large-scale environments. We found a predominant westward propagation of MCSs over NWSA, strongly influenced by easterly winds, but with some seasonal and regional variations likely related to ITCZ movement, moisture availability, topographic settings and vertical wind shear. Our results aim at having a representation of MCSs in NwSA in a climatological scale following a lagrangian approach. This study allows the identification of patterns like MCSs movement and growth rate, and predictable variables that can be a fundamental step for risk planning processes and to evaluate model simulations.