DESEMPENHO DE MODELOS SARIMA E SARIMAX PARA PREVISÃO DE VAZÕES MÉDIAS MENSAIS

Autores

  • José Rodrigues da Silva Neto jose.r.neto@ufv.br
    Instituto de Ciências Exatas e Tecnológicas, Universidade Federal de Viçosa, Rio Paranaíba, MG, Brasil. https://orcid.org/0009-0007-5560-4898
  • Frederico Carlos Martins de Menezes Filho frederico.menezes@ufv.br
    Instituto de Ciências Exatas e Tecnológicas, Universidade Federal de Viçosa, Rio Paranaíba, MG, Brasil. https://orcid.org/0000-0003-4874-0254

DOI:

https://doi.org/10.31413/nat.v14i3.21043


Palavras-chave:

modelagem hidrológica, modelos estocásticos, séries temporais

Resumo

A modelagem estocástica de séries temporais para a previsão de vazões tem se consolidado como uma alternativa relevante para o planejamento e a gestão de recursos hídricos em bacias hidrográficas não monitoradas. Este estudo teve como objetivo avaliar o desempenho dos modelos SARIMA e SARIMAX, considerando diferentes períodos de treino e de teste, bem como a influência da inclusão de variáveis exógenas, na previsão das vazões médias mensais de uma sub-bacia do rio São Francisco. A metodologia adotada seguiu a abordagem de Box e Jenkins (1976). Foram selecionados os modelos SARIMA (1,0,0)(4,0,0)[12] e SARIMAX (1,0,0)(2,0,0)[12], sendo este último incluindo a precipitação como variável exógena. As previsões foram realizadas para quatro horizontes (3, 6, 9 e 12 meses) em 2022. O desempenho foi avaliado por meio das métricas NSE, KGE e PBIAS. Em todos os critérios, o modelo SARIMAX superou o SARIMA. Verificou-se que a escolha dos intervalos de treino e de teste não foi determinante para a melhora do desempenho, sendo os melhores resultados obtidos com o maior período de treino (98% da série). O SARIMAX apresentou NSE = 0,65, KGE = 0,68 e PBIAS = 0,1% no treino, e NSE = 0,45, KGE = 0,5 e PBIAS = 20,9% no teste.

Palavras-chave: modelagem hidrológica; modelos estocásticos; séries temporais.

 

Performance of SARIMA and SARIMAX models for monthly mean streamflow forecasting

 

ABSTRACT: Stochastic time series modeling for streamflow forecasting has become an important approach for water resources planning and management in ungauged river basins. This study aimed to evaluate the performance of SARIMA and SARIMAX models under different training and testing periods, as well as the influence of including exogenous variables, in forecasting monthly mean streamflow in a sub-basin of the São Francisco River. The methodology followed the Box and Jenkins (1976) approach. The selected models were SARIMA (1,0,0)(4,0,0)[12] and SARIMAX (1,0,0)(2,0,0)[12], the latter including precipitation as an exogenous variable. Forecasts were generated for four forecasting horizons (3, 6, 9, and 12 months) during 2022. Model performance was evaluated using the Nash–Sutcliffe Efficiency (NSE), Kling–Gupta Efficiency (KGE), and Percent Bias (PBIAS) metrics. The SARIMAX model outperformed the SARIMA model across all evaluation criteria. The results indicated that the choice of training and testing intervals was not a determining factor in improving model performance, with the best results obtained using the longest training period (98% of the time series). During the training period, the SARIMAX model achieved an NSE of 0.65, KGE of 0.68, and PBIAS of 0.1%, whereas during the testing period, it yielded an NSE of 0.45, KGE of 0.50, and PBIAS of 20.9%.

Keywords: hydrological modeling; stochastic models; time series.

Referências

ALHARBI, F. R.; CSALA, D. A. Seasonal autoregressive integrated moving average with exogenous factors (SARIMAX) forecasting model-based time series approach. Inventions, v. 7, n. 4, e94, 2022. https://doi.org/10.3390/inventions7040094

AKAIKE, H. A new look at the statistical model identification. In: PARZEN, E.; TAMABE, K.; KITAGAWA, G. (Eds.). Springer Series in Sattistics. New York: Springer. 1974. p. 215-222. https://doi.org/10.1007/978-1-4612-1694-0_16

ANA_Agência Nacional de Águas e Saneamento Básico. Hidroweb: sistema de informações hidrológicas. Brasília, DF, 2024. Disponível em: https://www.snirh.gov.br/hidroweb/. Acesso em: 15 out. 2024.

BARBOSA, M. H.; CARVALHO, M. de M.; SÁFADI, T.; OLIVEIRA, L. F. C. de. Modelagem estocástica de séries temporais pluviométricas e fluviométrica da bacia hidrográfica do Rio Preto, Sudeste do Brasil. Revista Brasileira de Meteorologia, v. 39, e39240051, 2024. https://doi.org/10.1590/0102-77863910051

BARRIENTOS-TORRES, D.; MARTÍNEZ-RÍOS, E. A.; NAVARRO-TUCH, S. A.; PABLOS-HACH, J. L.; BUSTAMANTE-BELLO, R. Water flow modeling and forecast in a water branch of Mexico City through ARIMA and transfer function models for anomaly detection. Water, v. 15, n. 15, e2792, 2023. https://doi.org/10.3390/w15152792

BAYER, D. M.; CASTRO, N. M. dos R.; BAYER, F. M. Modelagem e previsão de vazões médias mensais do rio Potiribu utilizando modelos de séries temporais. Revista Brasileira de Recursos Hídricos, v. 17, n. 2, p. 229-239, 2012. https://doi.org/10.21168/rbrh.v17n2.p229-239

BAYER, F. M.; SOUZA, A. M. Forecasting with Wavelets and Traditional Models: A Comparative. Revista Brasileira de Biometria, v. 28, p. 40-61, 2010.

BLEIDORN, M. T.; PINTO, W. P.; BRAUM, E. S.; LIMA, G. B.; MONTEBELLER, C. A. Modelagem e previsão de vazões médias mensais do rio Jacu, ES, utilizando o modelo SARIMA. IRRIGA, v. 24, n. 2, p. 320-335, 2019. https://doi.org/10.15809/irriga.2019v24n2p320-335

BOX, G. E. P.; JENKINS, G. M. Time series analysis, forecasting and control. San Francisco: Holden-Day, 1976. 575p.

CAMINHA, A. R.; GUIMARÃES, L. R.; CORRÊA, F. V.; SOUZA, G. R. de; OLIVEIRA, L. F. C. de; SÁFADI, T. (2025). Previsão de precipitação e vazão a partir da modelagem de séries temporais. Revista Brasileira de Climatologia, v. 36, n. 21, p. 264-290, 2025. https://doi.org/10.55761/abclima.v36i21.18885

COSTA, G. E. de M. e; MENEZES FILHO, F. C. M. de; CANALES, F. A.; FAVA, M. C.; BRANDÃO, A. R. A.; PAES, R. P. Assessment of time series models for mean discharge modeling and forecasting in a sub-basin of the Paranaíba River, Brazil. Hydrology, v. 10, n. 11, e208, 2023. https://doi.org/10.3390/hydrology10110208

DAZZI, S.; VACONDIO, R.; MIGNOSA, P. Flood stage forecasting using machine-learning methods: a case study on the Parma River (Italy). Water, v. 13, n. 12, e1612, 2021. https://doi.org/10.3390/w13121612

DHANESH, Y.; BINDHU, V. M.; SENENT-APARICIO, J.; BRIGHENTI, T. M.; AYANA, E.; SMITHA, P. S.; FEI, C.; SRINIVASAN, R. A comparative evaluation of the performance of CHIRPS and CFSR data for different climate zones using the SWAT model. Remote Sensing, v. 12, e3088, 2020. https://doi.org/10.3390/rs12183088

DUC, L.; SAWADA, Y. A signal-processing-based interpretation of the Nash–Sutcliffe efficiency. Hydrology and Earth System Sciences, v. 27, n. 9, p. 1827-1839, 2023. https://doi.org/10.5194/hess-27-1827-2023

FAZLA, A.; AYDIN, M.E.; KOZAT, S.S. Joint optimization of linear and nonlinear models for sequential regression. Digital Signal Processing, v. 132, e103802, 2023. https://doi.org/10.1016/j.dsp.2022.103802

GUPTA, H. V.; SOROOSHIAN, S.; YAPO, P. O. Status of automatic calibration for hydrologic models: comparison with multilevel expert calibration. Journal of Hydrologic Engineering, v. 4, n. 2, p. 135-143, 1999. https://doi.org/10.1061/(ASCE)1084-0699(1999)4:2(135)

HARMEL, R. D.; SMITH, P. K.; MIGLIACCIO, K. W.; CHAUBEI, I.; DOUGLAS-MANKIN, K. R.; BENHAM, B.; SHUKLA, S.; MUÑOZ-CARPENA, R.; ROBSON, B. J. Evaluation, interpreting, and communicating performance of hydrologyc/water quality models considering intended use: A review and recommendations. Environmental Modelling & Software, v. 57, p. 40-51, 2014. https://doi.org/10.1016/j.envsoft.2014.02.013

HYNDMAN, R.J.; KHANDAKAR, Y. Automatic Time Series Forecasting: The Forecast Package for R. Journal of Statistical Software, v. 27, p. 1-22, 2008. https://doi.org/10.18637/jss.v027.i03

HYNDMAN, R. J.; ATHANASOPOULOS, G. Forecasting: principles and practice. 3. ed. Melbourne: OTexts, 2021. 442p.

KASSEM, A. A.; RAHEEM, A. M.; KHIDIR, K. M. Daily streamflow prediction for Khazir River Basin Using ARIMA and ANN Models. Zanco Journal of Pure and Applied Sciences, v. 32, p. 30-39, 2020. https://doi.org/10.21271/zjpas.32.3.4

KHODAKHAH, H.; AGHELPOUR, P.; HAMEDI, Z. Comparing linear and non-linear data-driven approaches in monthly river flow prediction, based on the models SARIMA, LSSVM, ANFIS, and GMDH. Environmental Science and Pollution Research, v. 29, p. 21935-21954, 2022. https://doi.org/10.1007/s11356-021-17443-0

KNOBEN, W. J. M; FREER, J. E.; WOODS, R. A. Inherent benchmark or not? Comparing Nash–Sutcliffe and Kling–Gupta efficiency scores. Hydrology and Earth System Sciences, v. 23, n. 10, p. 4323-4331, 2019. https://doi.org/10.5194/hess-23-4323-2019

MACHIWAL, D.; JHA, M. K. Hydrologic time series analysis: theory and practice. Dordrecht: Springer, 2012. 316p.

MENEZES FILHO, F. C. M. Modelagem e previsão de temperaturas médias mensais de Rio Paranaíba/MG utilizando modelos de séries temporais. Revista Ibero Americana de Ciências Ambientais, v. 11, n. 6, p. 251-261, 2020. http://doi.org/10.6008/CBPC2179-6858.2020.006.0021

MIAO, C.; GOU, J.; FU, B.; TANG, Q.; DUAN, Q.; CHEN, Z.; LEI, H.; CHEN, J.; GUO, J.; BORTHWICK, A. G. L.; DING, W.; DUAN, X.; LI, Y.; KONG, D.; GUO, X.; WU, J. High-quality reconstruction of China's natural streamflow. Science Bulletin, v. 67, p. 547-556, 2022. https://doi.org/10.1016/j.scib.2021.09.022

MORIASI, D. N.; GITAU, M. W.; PAI, N.; DAGGUPATI, P. Hydrologic and water quality models: performance measures and evaluation criteria. Transactions of the ASABE, v. 58, n. 6, p. 1763-1785, 2015. https://doi.org/10.13031/trans.58.10715

R CORE TEAM. R: a language and environment for statistical computing. Vienna: R Foundation for Statistical Computing, 2017.

SAID, S. E.; DICKEY, D. A. Testing for unit roots in autoregressive-moving average models of unknown order. Biometrika, v. 71, n. 3, p. 599-607, 1984. https://doi.org/10.2307/2336570

SHAH, H. S. M.; YASSIN, M. A.; ABBA, S. I.; LAWAL, D. U.; HUSSEIN AL-QADAMI, E. H.; TEO, F. Y.; MUSTAFFA, Z.; ALJUNDI, I. H. Flood risk and vulnerability from a changing climate perspective: an overview focusing on flash floods and associated hazards in Jeddah. Water, v. 15, n. 20, e3641, 2023. https://doi.org/10.3390/w15203641

SIVAKUMAR, M.; PARTHASARATHY, S.; PADMAPRIYA, T. Trade-off between training and testing ratio in machine learning for medical image processing. PeerJ Computer Science, v. 10, e2245, 2024. https://doi.org/10.7717/peerj-cs.2245

SUN, Y.; NIU, J.; SIVAKUMAR, B. A. Comparative Study of Models for Short-Term Streamflow Forecasting with Emphasis on Wavelet-Based Approach. Stochastic Environmental Research and Risk Assessment, v. 33, p. 1875-1891, 2019. https://doi.org/10.1007/s00477-019-01734-7

Downloads

Publicado

2026-08-13

Edição

Seção

Ciências Ambientais / Environmental Sciences

Como Citar

DESEMPENHO DE MODELOS SARIMA E SARIMAX PARA PREVISÃO DE VAZÕES MÉDIAS MENSAIS. (2026). Nativa, 14(3), e21043. https://doi.org/10.31413/nat.v14i3.21043