Development and Operational Testing of a Super‐Ensemble Artificial Intelligence Flood‐Forecast Model for a Pacific Northwest River

Coastal catchments in British Columbia, Canada, experience a complex mixture of rainfall- and snowmelt-driven contributions to flood events. Few operational flood-forecast models are available in the region. Here, we integrated a number of proven technologies in a novel way to produce a super-ensemble forecast system for the Englishman River, a flood-prone stream on Vancouver Island. This three-day-ahead modeling system utilizes up to 42 numerical weather prediction model outputs from the North American Ensemble Forecast System, combined with six artificial neural network-based streamflow models representing various slightly different system conceptualizations, all of which were trained exclusively on historical high-flow data. As such, the system combines relatively low model development times and costs with the generation of fully probabilistic forecasts reflecting uncertainty in the simulation of both atmospheric and terrestrial hydrologic dynamics. Results from operational testing by British Columbia's flood forecasting agency during the 2013-2014 storm season suggest that the prediction system is operationally useful and robust.

[1]  P. Coulibaly,et al.  Two decades of anarchy? Emerging themes and outstanding challenges for neural network river forecasting , 2012 .

[2]  S. Sorooshian,et al.  Multi-model ensemble hydrologic prediction using Bayesian model averaging , 2007 .

[3]  Nicholas Kouwen,et al.  Integrating Logistical and Technical Criteria into a Multiteam, Competitive Watershed Model Ranking Procedure , 2013 .

[4]  Michael D. Dettinger,et al.  North–South Precipitation Patterns in Western North America on Interannual-to-Decadal Timescales , 1998 .

[5]  C. Sivapragasam,et al.  Genetic programming model for forecast of short and noisy data , 2007 .

[6]  R. Guthrie,et al.  The Vulnerability of Lower Englishman River to Modelled Climate Change , 2003 .

[7]  Paul H. Whitfield,et al.  Spatiotemporal mapping of ENSO and PDO surface meteorological signals in British Columbia, Yukon, and southeast Alaska , 2010 .

[8]  Sungwon Kim,et al.  Flood Forecasting Using Neural Computing Techniques and Conceptual Class Segregation , 2013 .

[9]  William W. Hsieh Machine Learning Methods in the Environmental Sciences: Contents , 2009 .

[10]  E. Roulin,et al.  Skill and relative economic value of medium-range hydrological ensemble predictions , 2006 .

[11]  Roland B. Stull,et al.  Hydrometeorological Accuracy Enhancement via Postprocessing of Numerical Weather Forecasts in Complex Terrain , 2008 .

[12]  Sean W. Fleming An Informal Survey of Watershed Model Users: Preferences, Applications, and Rationales , 2009 .

[13]  Roland B. Stull,et al.  Bias-corrected short-range Member-to-Member ensemble forecasts of reservoir inflow , 2013 .

[14]  Sean W. Fleming,et al.  Streamflow Modelling: A Primer on Applications, Approaches and Challenges , 2012 .

[15]  Alex J. Cannon,et al.  Daily streamflow forecasting by machine learning methods with weather and climate inputs , 2012 .

[16]  E. Lorenz Deterministic nonperiodic flow , 1963 .

[17]  Honggang Zhang,et al.  A reservoir flood forecasting and control system for China / Un système chinois de prévision et de contrôle de crue en barrage , 2004 .

[18]  S. Sorooshian,et al.  Multimodel Combination Techniques for Analysis of Hydrological Simulations: Application to Distributed Model Intercomparison Project Results , 2006 .

[19]  Darrell G. Fontane,et al.  Spatial Decision Support System for Integrated River Basin Flood Control , 2002 .

[20]  A. Tokar,et al.  Rainfall-Runoff Modeling Using Artificial Neural Networks , 1999 .

[21]  L. Vasiliades,et al.  Flood producing mechanisms identification in southern British Columbia, Canada , 2000 .

[22]  P. Whitfield,et al.  Regime‐dependent streamflow sensitivities to Pacific climate modes cross the Georgia–Puget transboundary ecoregion , 2007 .

[23]  V. Singh,et al.  Mathematical Modeling of Watershed Hydrology , 2002 .

[24]  J. Nash,et al.  River flow forecasting through conceptual models part I — A discussion of principles☆ , 1970 .

[25]  Barbara G. Brown,et al.  Forecast verification: current status and future directions , 2008 .

[26]  Dong-Jun Seo,et al.  Towards the characterization of streamflow simulation uncertainty through multimodel ensembles , 2004 .

[27]  Tirusew Asefa,et al.  Ensemble Streamflow Forecast: A GLUE‐Based Neural Network Approach 1 , 2009 .

[28]  András Bárdossy,et al.  Generic error model for calibration and uncertainty estimation of hydrological models , 2008 .

[29]  Mohammad Bagher Menhaj,et al.  Training feedforward networks with the Marquardt algorithm , 1994, IEEE Trans. Neural Networks.

[30]  M. Castellano-Méndez,et al.  Modelling of the monthly and daily behaviour of the runoff of the Xallas river using Box-Jenkins and neural networks methods , 2004 .