Uso de padrões atmosféricos climatológicos na previsão diária de eventos de precipitação intensa no Rio Grande do Sul

• Autor(a) principal: Cebalhos, Vinícius de Souza
• Data de Publicação: 2020
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Manancial - Repositório Digital da UFSM
• dARK ID: ark:/26339/001300000x9vs
• Texto Completo: http://repositorio.ufsm.br/handle/1/26249
• topic: Precipitação extrema; Previsão por padrões; Padrões atmosféricos; CNPQ::CIENCIAS EXATAS E DA TERRA::GEOCIENCIAS::METEOROLOGIA

This work aims to carry out a quantitative and qualitative comparative assessment of the atmospheric patterns presented by Santos (2012) with the output of weather forecasting models, in order to predict the location of the occurrence of rain events with greater accuracy. For this, 3 precipitation events were selected in which there was an accumulation greater than 50 mm in 24 hours. The forecast data obtained for comparisons with the Santos (2012) patterns were the GFS (Global Forecast System) forecasts, using their daily departures from 12 UTC to 48 hours of forecast. GFS data was stored from January 2017 to June 2019, in the form of a database. The variables used were pressure at mean sea level, layer thickness between 500 hPa and 1000 hPa, geopotential height at 500 hPa, zonal wind component at 850 hPa, southern wind component at 850 hPa, specific humidity at 850 hPa, temperature air at 850 hPa, zonal wind component at 200 hPa and southern wind component at 200 hPa. The simulations obtained and the Santos (2012) patterns , were converted to matrices and thus applied to Pearson’s correlation test, to verify the quantitative similarity of the GFS simulations with the 5 Santos (2012) patterns . Event 1, registered on July 20, 2018, was identified as a cold front, associated with a cyclogenesis over the Atlantic Ocean and showing greater similarity with Santos (2012) Pattern 3 . Quantitatively, a similarity with Pattern 3 was also identified, observing the mean sea level in pressure, a variable that showed a clearer behavior in differentiating the forecast with the patterns, a correlation of 0.84 in the forecast 24 hours before the event and 0.854 in forecast 48 hours before the event. Event 2, recorded on July 25, 2018, is due to a suptropical Cyclonic Vortex in High Levels, resulting from the breaking of the excavated axis at 500 hPa. Qualitatively, this event is similar to Pattern 2, but quantitatively it was not possible to identify a higher correlation index. Event 3, observed an mesoscale convective system (MCS), with characteristics of a prefrontal with a 500 hPa dug over the Andes, low thickness gradient and intensification of pressure over southeastern South America, are similar to Santos (2012) pattern 1 on November 17, 2018, showed a correlation of 0.8412 in the 24-hour simulation and 0.783 in the simulation 48 hours before the event. Thus, perhaps to the limited number of cases analyzed in this study, the recognition was satisfactory for the cold front and MCS patterns, since these are the main precipitating systems in Rio Grande do Sul.