We found a match
Your institution may have rights to this item. Sign in to continue.
- Title
Previsão de Níveis Estáticos de Água Subterrânea em Poço de Monitoramento Utilizando Metodologia Estocástica.
- Authors
Boeno Patricio Luiz, Thiago; Schröder, Thomas; Silvério da Silva, José Luiz
- Abstract
This study aims to model and analyze the natural behavior of static level of a time series of groundwater level and precipitation in order to make forecast the levels for 6 months horizon. For this purpose, a stochastic model of ARMA class (autoregressive and moving average) was applied in 7-day interval data of groundwater static level and rainfall. The geological environment of the well is characterized by sedimentary formations of the Paraná Basin border, located in Alegrete municipality of the Rio Grande do Sul State. The monitoring well is equipped with automatic data logger and the time series under consideration was over 2 years long without failed measurements. Autoregressive and moving average models were tested until the third order of autocorrelation. The best model was selected based on AIC (Akaike Information Criterion) and BIC (Bayesian Information Criterion) values, where the chosen model was the ARMA (2, 0) which obtained the best fit. The error analysis allowed to observe the good performance of the model based on the graph of standardized residuals, attesting the predictive quality of chosen model. The evaluation of model coefficients allowed the establishment of an approximate relationship in millimeters of 0.31: 1 for the variables static level and rainfall respectively.
- Publication
Anuario do Instituto de Geociencias, 2017, Vol 40, Issue 2, p53
- ISSN
0101-9759
- Publication type
Article
- DOI
10.11137/2017_2_53_60