Brazilian electrical storms tracking and forecast system
Mariana Kleina1; Eduardo Alvim Leite2
1 Universidade Federal do Paraná - Departamento de Engenharia de Produção; 2 Sistema Meteorólogico do Paraná
doi:10.20906/CPS/CILAMCE2017-0966
Resumo
In this work, we aim to perform monitoring and forecast of a set of lightnings through space and time (called electrical storms) for all the brazilian territory in real time using only lightning data (no other remote sensing means). The Probabilistic Forecast of Lightning Events System (SIPPER) proposed here does the tracking for up to three hours of the storm core path minutely, delimiting the area where the storm develops. Moreover, this system also predicts the storm behavior for the next hour, pointing out which areas will most likely be affected and which ones won't, characterized by and elliptical representation of lightning incidence. Some of the variables provided by the system are: lowest and highest lightning occurrence probabilities, geographic coordinates of the storm core represented by it latitude and longitude, density (expected number of lightnings by square kilometer), direction and speed of displacement and occurrence of high electrical peak current. The Convergent Data Sharpening method is used to develop the monitoring part of the system, and extrapolation methods are used for the forecast, all through the R software environment. A platform for data visualization was also created to observe by the minute storms displacement and forecast all over Brazil. Through geoprocessing technologies, the system visualization is possible in different visualisation layers, such as cities, states, highways, transmission lines and electrical energy distribution nets. The lightning data that feeds the system come from the National Network of Atmospheric Discharges Detection (Rindat) and from the GLD360 service (Global Lightning Dataset) provided by the finnish company Vaisala. The first source covers the South, Southeast and Midwest of the country, and the later one covers all Brazil and some neighbor countries, including additionally sea areas. This way, the system process in real time lightning data for the national territory, sometimes processing a hundred thousand lightnings. As a result, it is possible to
Palavras-chave: Electrical storms; Tracking; Forecast; Scientific visualization