ABSTRACT
This paper describes a dataset of convective systems (CSs) associated with hailstorms over Brazil tracked using GOES-16 Advanced Baseline Imager (ABI) measurements and the Tracking and Analysis of Thunderstorms (TATHU) tool. The dataset spans from June 5, 2018, to September 30, 2023, providing five-year period of storm activity. CSs were detected and tracked using the ABI's clean IR window brightness temperature at 10.3 µm, projected on a 2 km x 2 km Lat-Lon WGS84 grid. Systems were identified using a brightness temperature (BT) threshold of 235 K, conducive to detecting convective clusters with larger area and excluding smaller or non-convective cells such as groups of thin Cirrus clouds. Each detected CS was treated as an object, containing geographic boundaries and raster statistics such as BT's mean, minimum, standard deviation, and count of data points within the CS polygon, which serves as proxy for size estimates. The life cycle of each system was tracked based on a 10 % overlap area criterion, ensuring continuity, unless disrupted by dissociative or associative events. Then, the tracked CSs were filtered for intersections in space and time with verified ground reports of hail, from the Prevots group. The matches were then exported to a database with SpatiaLite enabled data format to facilitate spatial data queries and analyses. This database is structured to support advanced research in severe weather events, in particular hailfall. This setting allows for extensive temporal and spatial analyses of convective systems, making it useful for meteorologists, climate scientists, and researchers in related fields . The inclusion of detailed tracking information and raster statistics offers potential for diverse applications, including climate model validation, weather prediction enhancements, and studies on the climatological impact of severe weather phenomena in Brazil.
ABSTRACT
In this work, we introduce a formalism to highlight the role of decision-making implicit in the setup of early warning systems (EWSs) and its consequences with respect to loss avoidance for end users. The formalism, a close relative of the cost/loss approach, combines EWS verification scores with traditional expressions of risk from the point of view of the user. This formalism articulates in mathematical format many well-known issues surrounding EWS usage, offering a conceptual anchor for concepts that otherwise may seem to wobble among the multidisciplinary perspectives participating in the EWS chain. This decision model is visually represented in a variation of the popular "performance diagram" used in forecast and warning verification. Our diagram adds to this the perspective of a generic user, in an effort to gain insight into how choices made regarding EWS settings may determine which users benefit from warnings and which do not. Although these results are based on a conceptual model, they are useful to better understand the actual benefits experienced by users and to highlight aspects that may temper unrealistic expectations on EWSs. The recent United Nations initiative to extend EWSs for natural hazards to all nations within 5 years will make EWSs more common and more public. The approach proposed here can be a tool to promote greater transparency and improve the necessary dialog between warning issuers and users in order to reduce loss.