Com qual antecedência conseguimos prever cheias no rio Uruguai usando um modelo hidrológico de grande escala?

• Autor(a) principal: GUIMARÃES, Guilherme Mendoza
• Data de Publicação: 2018
• Outros Autores: FAN, Fernando Mainardi, MARCUZZO, Francisco Fernando Noronha, BUFFON, Franco Turco, GERMANO, Andrea de Oliveira
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Repositório Institucional de Geociências - RIGEO
• Texto Completo: https://rigeo.sgb.gov.br/handle/doc/19640
• topic: HIDROLOGIA; MODELOS HIDROLÓGICOS; SISTEMA DE ALERTA; PREVISÃO HIDROLÓGICA; PREVISIBILIDADE; RIO GRANDE DO SUL; RIO URUGUAI; SISTEMA DE ALERTA DE EVENTOS CRÍTICOS; SACE; CHEIAS; INUNDAÇÕES; MODELOS DE GRANDES BACIAS; MGB; MODELO DIGITAL DE ELEVAÇÃO; OTTOCODIFICATION; RIVER CODIFICATION; MORPHOLOGY; RELIEF; HIPSOMETRIC; DIGITAL MODEL ELEVATION; HIDROLOGY; HYDROLOGICAL MODELS; ALERT SYSTMS; HYDROLOGICAL FORECAST; URUGUAY RIVER; BRASIL; ARGENTINA; URUGUAI; PREVISIÓN HIDROLÓGICA; HIDROLOGÍA

The short and medium-range flow forecasting techniques using large-scale hydrological models have several direct applications in the management of natural disasters through warning systems. The present study is a research study on this type of system, applied in the Uruguay River Basin (RS, SC, Argentina and Uruguay). The objective of this study was to investigate the predictability of critical events at points of interest in the Uruguay river basin. We aimed to evaluate how early it is possible to estimate the peak flow in some sites susceptible to flooding in the basin. For the present research we selected municipalities that will be initially served by the Geological Service of Brazil (CPRM) warning system in the Uruguay river: Garruchos, Itaqui, Porto Lucena, São Borja and Uruguaiana. In the evaluation, the Nash-Sutcliffe Efficiency coefficient (NSE) was used. After the calibration and validation of the model, the predictability analysis was performed based on 15 flood events that occurred between 1980 and 2017 in which the forecasts of daily time-step were compared to a reference simulation. It was verified that there is an increase in the predictability from one day for up to three days, increasing as further downstream the place of interest is located. Also it was evidenced that the predictability especially in Uruguaiana city is dependent on where the floods are originated (higher Uruguay basin or Ibicuí basin). These results constitute useful information that may assist managers in decision making during critical events where forecast are provided by the proposed type of modeling