Forecasting daily river flows using nonlinear time series models

Summary In the hydrology studies, it is well known that the river flows are affected by various factors, and therefore the dynamics in their associated time series are complicated and have nonlinear behaviors. In an empirical study, we investigate the capability of five classes of nonlinear time series models, namely Threshold Autoregressive (TAR), Smooth Transition Autoregressive (STAR), Exponential Autoregressive (EXPAR), Bilinear Model (BL) and Markov Switching Autoregressive (MSAR) to capture the dynamics in the Colorado river discharge time series. Least Squares (LS) and Maximum Likelihood (ML) methods are employed to estimate parameters of the models. For model comparison three criteria, namely loglikelihood, Akaike information criterion (AIC) and Bayesian information criterion (BIC) are calculated. The results show that a self-exciting TAR (SETAR) model performs better than other four competing models. To forecast future river discharge values an iterative method is applied and forecasting confidence intervals are constructed. The out-of-sample 1-day to 5-day ahead forecasting performances of the models based on ten forecast accuracy measures are evaluated. Comparing verification metrics of all models, SETAR model presents the best forecasting performance.

[1]  Jianqing Fan,et al.  Nonlinear Time Series : Nonparametric and Parametric Methods , 2005 .

[2]  G. Schwarz.  Estimating the Dimension of a Model , 1978 .

[3]  Ronald R. Mohler,et al.  Natural Bilinear Control Processes , 1970, IEEE Trans. Syst. Sci. Cybern..

[4]  Philip Hans Franses,et al.  A seasonal periodic long memory model for monthly river flows , 1998, Environ. Model. Softw..

[5]  L. Márkus,et al.  A long range dependent model with nonlinear innovations for simulating daily river flows , 2004 .

[6]  Ozgur Kisi,et al.  Wavelet regression model as an alternative to neural networks for monthly streamflow forecasting , 2009 .

[7]  Ruey S. Tsay,et al.  Analysis of Financial Time Series , 2005 .

[8]  Shlomo Havlin,et al.  Multifractality of river runoff and precipitation: Comparison of fluctuation analysis and wavelet methods , 2003 .

[9]  Wen Wang.  Stochasticity, nonlinearity and forecasting of streamflow processes , 2006 .

[10]  Bruce E. Hansen,et al.  Testing for linearity , 1999 .

[11]  James G. MacKinnon,et al.  Heteroskedastcity-robust tests in regressions directions , 1985 .

[12]  H. White.  A Heteroskedasticity-Consistent Covariance Matrix Estimator and a Direct Test for Heteroskedasticity , 1980 .

[13]  H. Tong,et al.  Threshold Autoregression, Limit Cycles and Cyclical Data , 1980 .

[14]  B. Mandelbrot.  Limit Theorem on the Self-Normalized Range for Weakly and Strongly Dependent Process , 1975 .

[15]  C. Granger,et al.  Modelling Nonlinear Economic Relationships , 1995 .

[16]  D. Legates,et al.  Evaluating the use of “goodness‐of‐fit” Measures in hydrologic and hydroclimatic model validation , 1999 .

[17]  H. Tong.  A Personal Overview Of Nonlinear Time-Series Analysis From A Chaos Perspective , 1995 .

[18]  A Bunde,et al.  Comment on "Scaling of atmosphere and ocean temperature correlations in observations and climate models". , 2004, Physical review letters.

[19]  Timo Teräsvirta,et al.  Aspects of modelling nonlinear time series , 1986 .

[20]  Comparing linear and bilinear models on water level for the Kelantan River in Malaysia , 2005 .

[21]  D. Schertzer,et al.  Multifractal analysis of daily river flows including extremes for basins of five to two million square kilometres, one day to 75 years , 1998 .

[22]  Keith W. Hipel,et al.  Advances in Box‐Jenkins modeling: 2. Applications , 1977 .

[23]  Absolute and relative measures for evaluating the forecasting performance of time series models for daily streamflows , 2006 .

[24]  B. LeBaron,et al.  A test for independence based on the correlation dimension , 1996 .

[25]  Jon D. Pelletier,et al.  Long-range persistence in climatological and hydrological time series: analysis, modeling and application to drought hazard assessment , 1997 .

[26]  Jean-Yves Pitarakis,et al.  Estimation and Model Selection Based Inference in Single and Multiple Threshold Models , 2002 .

[27]  J. Amorocho,et al.  The nonlinear prediction problem in the study of the runoff cycle , 1967 .

[28]  Ruey S. Tsay,et al.  Testing and Modeling Threshold Autoregressive Processes , 1989 .

[29]  Michael P. Clements,et al.  The Performance of Alternative Forecasting Methods for SETAR Models , 1997 .

[30]  Péter Elek,et al.  A two-state regime switching autoregressive model with an application to river flow analysis , 2007 .

[31]  H. Akaike.  A new look at the statistical model identification , 1974 .

[32]  Jiahui Wang,et al.  Modeling Financial Time Series with S-PLUS® , 2003 .

[33]  Murad S. Taqqu,et al.  A seasonal fractional ARIMA Model applied to the Nile River monthly flows at Aswan , 2000 .

[34]  A. I. McLeod,et al.  ARMA MODELLING WITH NON-GAUSSIAN INNOVATIONS , 1988 .

[35]  P. Hubert,et al.  Multifractal analysis and modeling of rainfall and river flows and scaling, causal transfer functions , 1996 .

[36]  Shlomo Havlin,et al.  Long-term persistence and multifractality of river runoff records: Detrended fluctuation studies , 2003 .

[37]  Marcella Niglio,et al.  Multi-step SETARMA predictors in the analysis of hydrological time series , 2006 .

[38]  Paulin Coulibaly,et al.  Nonstationary hydrological time series forecasting using nonlinear dynamic methods , 2005 .

[39]  John S. Gulliver,et al.  Prediction of River Discharge at Ungaged Sites with Analysis of Uncertainty , 1993 .

[40]  R. Abrahart,et al.  Comparing neural network and autoregressive moving average techniques for the provision of continuous river flow forecasts in two contrasting catchments , 2000 .

[41]  Klaus Fraedrich,et al.  Scaling of atmosphere and ocean temperature correlations in observations and climate models. , 2003, Physical review letters.

[42]  William A. Brock,et al.  8 Nonlinear time series and macroeconometrics , 1993 .

[43]  Timo Teräsvirta,et al.  Testing linearity against smooth transition autoregressive models , 1988 .

[44]  Clive W. J. Granger,et al.  OVERVIEW OF NONLINEAR MACROECONOMETRIC EMPIRICAL MODELS , 2001, Macroeconomic Dynamics.

[45]  J. Amorocho,et al.  Determination of Nonlinear Functional Response Functions in Rainfall-Runoff Processes , 1971 .

[46]  Timo Teräsvirta,et al.  Smooth transition autoregressive models - A survey of recent developments , 2000 .

[47]  Michael P. Clements,et al.  A Comparison of the Forecast Performance of Markov�?Switching and Threshold Autoregressive Models of Us Gnp , 1998 .

[48]  Philip Hans Franses,et al.  Non-Linear Time Series Models in Empirical Finance , 2000 .

[49]  R. Tsay.  Testing and modeling multivariate threshold models , 1998 .

[50]  Heung Wong,et al.  Streamflow forecasting using functional-coefficient time series model with periodic variation. , 2009 .

[51]  Norman R. Swanson,et al.  Nonlinear Econometric Modelling: A Selective Review , 1999 .

[52]  T. Ozaki,et al.  Modelling nonlinear random vibrations using an amplitude-dependent autoregressive time series model , 1981 .

[53]  H. Tong.  On a threshold model , 1978 .

[54]  T. Rao.  On the Theory of Bilinear Time Series Models , 1981 .

[55]  P. Schmidt,et al.  On the Power of the KPSS Test of Stationarity Against Fractionally-Integrated Alternatives , 1996 .

[56]  H. Tong.  Non-linear time series. A dynamical system approach , 1990 .

[57]  M. Medeiros,et al.  Linear models, smooth transition autoregressions, and neural networks for forecasting macroeconomic time series: A re-examination , 2005 .

[58]  S. L. Jacoby.  A mathematical model for nonlinear hydrologic systems , 1966 .

[59]  M. Taqqu,et al.  Fractionally differenced ARIMA models applied to hydrologic time series: Identification, estimation, and simulation , 1997 .

[60]  Radko Mesiar,et al.  Comparison of forecasting performance of nonlinear models of hydrological time series , 2006 .

[61]  Douglas W. Nychka,et al.  A personal overview of non-linear time series analysis from a chaos perspective. Commentary , 1995 .

[62]  Fallaw Sowell.  Maximum likelihood estimation of stationary univariate fractionally integrated time series models , 1992 .

[63]  Peter J. Brockwell,et al.  On the general bilinear time series model , 1988 .

[64]  Kung-Sik Chan,et al.  Time Series Analysis: With Applications in R , 2010 .

[65]  A. I. McLeod,et al.  Preservation of the rescaled adjusted range: 1. A reassessment of the Hurst Phenomenon , 1978 .

[66]  Xian-Huan Wen,et al.  A Nonlinear Time Series Model for Spring Flow: An Example from Shanxi Province, China , 1998 .

[67]  R. Vogel,et al.  Probability Distribution of Low Streamflow Series in the United States , 2002 .

[68]  W. E. Watt,et al.  Nested threshold autoregressive (NeTAR) models , 1997 .

[69]  Philip Hans Franses,et al.  Modeling Multiple Regimes in the Business Cycle , 1999, Macroeconomic Dynamics.

[70]  James D. Hamilton.  A New Approach to the Economic Analysis of Nonstationary Time Series and the Business Cycle , 1989 .

[71]  Lanh Tat Tran,et al.  On the first-order bilinear time series model , 1981, Journal of Applied Probability.

[72]  C. Granger,et al.  An introduction to bilinear time series models , 1979 .

[73]  A. Raftery,et al.  Space-time modeling with long-memory dependence: assessing Ireland's wind-power resource. Technical report , 1987 .

[74]  Simon M. Potter.  Nonlinear Time Series Modelling: An Introduction , 1999 .

[75]  H. E. Hurst,et al.  Long-Term Storage Capacity of Reservoirs , 1951 .

[76]  T. Rao,et al.  An Introduction to Bispectral Analysis and Bilinear Time Series Models , 1984 .

[77]  T. Teräsvirta.  Specification, Estimation, and Evaluation of Smooth Transition Autoregressive Models , 1994 .