Rainfall in the Amazon Basin is very heterogeneous, mainly because the area encompassed is quite large. Among the systems responsible for rainfall, some stand out as extreme storm events. This study presents a criterion for identifying potentially severe convection (PSC) in the Amazon region from the Tropical Rainfall Measuring Mission (TRMM) database, specifically from Precipitation Features (PF)-1998-2012-generated and stored by the University of Utah. The seasonal and spatial distributions are similar to distributions of Mesoscale Convective Systems already cataloged in previous studies based on GOES satellite images. The seasons with the highest number of cases are austral spring, winter, and fall. With the Amazon region divided into six subregions and cases accumulated by quarter (JFM, AMJ, JAS, OND) the south of the Amazon subregion (SA) accounts for the largest number of cases with the OND quarter with higher occurrence and the lowest in AMJ. Different diurnal cycles of PSC are observed across the region with the more western areas, closer to the Andes, favoring nighttime cases, especially in the austral spring and summer. The diurnal cycle of the number of the most extreme cases is more pronounced than the diurnal cycle when a large collection of deep convection cases are included.

We describe the existence of an Amazonian low-level jet (ALLJ) that can affect the propagation and life cycle of convective systems from the northeast coast of South America into central Amazonia. Horizontal winds from reanalysis were analyzed during March, April, and May (MAM) of the two years (2014–2015) of the GoAmazon2014/5 field campaign. Convective system tracking was performed using GOES-13 infrared imagery and classified into days with high and weak convective activity. TheMAMaverage winds show a nocturnal enhancement of low-level winds starting near the coast in the early evening and reaching 1600 km inland by late morning. Mean three-hourly wind speeds maximize at 9-10 m/s near 900 hPa, but individual days can have nighttime low-level winds exceeding 12 m/s. Based on objective low-level wind criteria, the ALLJ is present 10-40% of the time over the Amazon during MAM depending on the location and time of day. The evolution of the ALLJ across the Amazon impacts the frequency of occurrence of cloud clusters and the intensity of the moisture flux. In addition, the ALLJ is associated with the enhancement of northeasterly flow in the mid-troposphere during active convective days, when vertical momentum transport may be occurring in the organized cloud clusters. During the weakly active convective period, the ALLJ is weaker near the coast but stronger across the central Amazon and appears to be linked more directly with the South American low-level jet.

Convective system tracking was performed using 30‐min GOES‐13 infrared imagery over the Amazon region during 2014 and 2015. A total of 116,701 convective systems were identified and statistics on the probability of occurrence of track area, lifetime, and system velocity were analysed. Maps of the total and seasonal geographic density of trajectories and the geographic density of clusters at genesis, during propagation, and at dissipation were also assessed. The mean area and lifetime of the tracked systems was 4 × 104 km2 and 3 hr, respectively. The top 10% largest systems had areas >8 × 104 km2 and the top 10% longest lived systems lasted >7 hr. The geographical distribution of clusters identified on the coast and within the Amazon basin varied seasonally and their life cycle tracking showed that they are typically distinct from one another (i.e., it is relatively rare for systems to start at the coast and propagate 1,500 km to the centre of the basin). Although the average system velocity indicated a predominantly westward motion, a large spread in the direction of propagation was found. In particular, the probability of a meridional component of motion was generally the same for northward or southward directions and 35% of the zonal propagation was associated with eastward movement. The presence of Kelvin waves accounted for some of the eastward system motion, in addition to increasing the area and lifetime of storms compared to when Kelvin waves were not present.

This study presents a slow antenna that has been develop to work as a lightning warning system (LWS). The sensor has been developed by STORM-T Laboratory of the University of São Paulo as part of a research and development program of the Companhia Energética do Ceará (COELCE). The LWS built can detect intra-cloud (IC) and cloud to ground (CG) lightning that is used to predict the probability of CG occurrence. The system is a slow-antenna integrated with a microcontroller and a single board computer that proportionate a compact design and low cost. Furthermore, the firmware can be update to include more detecting features anytime.