Voltar
Artigo Acesso aberto Produção Nacional
BIBTEX RIS

Modelagem e Previsão de Vazões Médias Mensais do Rio Potiribu Utilizando Modelos de Séries Temporais

2012; Brazilian Association of Water Resources; Volume: 17; Issue: 2 Linguagem: Inglês

10.21168/rbrh.v17n2.p229-239

ISSN

2318-0331

Autores

Débora Bayer, Nilza Maria dos Reis Castro, Fábio M. Bayer,

Tópico(s)

Hydrology and Drought Analysis

Resumo

This paper considers analyzing and modeling a time series of mean monthly streamflows using the stochastic model ARIMA (autoregressive integrated moving average). This analysis aims to model and forecast the monthly streamflow in the Potiribu river basin, located in the northwest region of Rio Grande do Sul state, Brazil. We tested several models of the SARIMA type, or ARIMA models which take the seasonal characteristics of the data into account. Among all tested models we selected some competing ones that had the lowest Akaike information criteria (AIC). Finally, we chose the SARIMA model (3,0,0) (2,1,2)12 which presented the best fit to the series of flows observed, with the Nash-Sutclife coefficient (NSC) equal to 0.68, as training data. The model was used to make forecasts for the six months forward, resulting in NSC = 0.81 as validating data

Referência(s)