Abstract

RIVERO-VILLAVERDE, Anaily; ALONSO-BRITO, Gustavo Reinel  and  LAU-QUAN, Andrés. Stochastic Linear Modeling for the Forecast of Flows in Basin, Western Region of Cuba. Rev Cie Téc Agr [online]. 2018, vol.27, n.4 ISSN 2071-0054.

This study focuses on the subwatershed "V Aniversario", belonging to Cuyaguateje River Basin, Pinar del Río Province. Its latitudinal position leads to the development of a greater volume of runoff, higher than other karstic regions in the country, then numerous damages related mainly with floods are done. Therefore, the objective of this work is to forecast the runoff in advance at the subwatershed, on annual and monthly scales, through linear stochastic modeling. In order to comply with the above, White Noise, AR (p), MA (q) and ARMA (p, q) models for the annual runoff series were implemented in R software, being the White Noise model the best adjusted to it. However, the monthly runoff series must be modeled by SARIMA, because it presented a seasonal behavior. It is interesting to note that the latter had a better memory and linear correlation compared to the annual series.

Keywords : time series; White Noise; AR (p); MA (q); autocorrelation.

        · abstract in Spanish     · text in English | Spanish     · English ( pdf ) | Spanish ( pdf )