Specifying a hierarchical mixture of experts for hydrologic modeling: Gating function variable selection

[1] The complexity of predicting surface runoff from hydrological models is compounded by uncertainties associated with the model structure, parameters and inputs. A hierarchical mixture of experts (HME) is recognized as one of the ways of incorporating model structural uncertainty into hydrological simulations. In this article, a framework capable of incorporating parameter and structural uncertainties is implemented via the use of a hierarchical mixture of experts together with sequential Monte Carlo parameter sampling. The use of a HME enables aggregation of multiple constituent models at the same instance, mixed to different extents in a dynamic manner as specified by a gating function, allowing the modeler to better characterize the uncertainty associated with the obtained predictions. This article presents a mechanism for better specifying the structure of the gating function used for combining models in a HME approach, by investigating the combination of predictor variables that allows the best model mixing. These predictors exist in various forms, each of which represents information on the catchment. We apply three different types of predictors to a case study, the Never Never River catchment in Australia. The outcomes from this case study consistently demonstrate improved Bayesian information criterion (BIC) readings for the HME especially when used with a combination of predictors. The predictor coefficients are further used for regionalization with the Manning River catchment, having similar characteristics to the Never Never River catchment, and also demonstrate satisfactory improvement in BIC when compared with a single structure model.

[1]  Ashish Sharma,et al.  Multisite seasonal forecast of arid river flows using a dynamic model combination approach , 2009 .

[2]  Mark S. Gordon,et al.  Chlorine activation indoors and outdoors via surface-mediated reactions of nitrogen oxides with hydrogen chloride , 2009, Proceedings of the National Academy of Sciences.

[3]  D. S. Young,et al.  Mixtures of regressions with predictor-dependent mixing proportions , 2010, Comput. Stat. Data Anal..

[4]  K. P. Sudheer,et al.  Effect of spatial resolution on regionalization of hydrological model parameters , 2012 .

[5]  George Kuczera,et al.  Combining site and regional flood information using a Bayesian Monte Carlo approach , 2009 .

[6]  Peter Reichert,et al.  Analyzing input and structural uncertainty of nonlinear dynamic models with stochastic, time‐dependent parameters , 2009 .

[7]  Nando de Freitas,et al.  Sequential Monte Carlo Methods in Practice , 2001, Statistics for Engineering and Information Science.

[8]  Thorsten Wagener,et al.  Convergence of approaches toward reducing uncertainty in predictions in ungauged basins , 2011 .

[9]  Thorsten Wagener,et al.  Parameter estimation and regionalization for continuous rainfall-runoff models including uncertainty , 2006 .

[10]  A. Raftery,et al.  Using Bayesian Model Averaging to Calibrate Forecast Ensembles , 2005 .

[11]  Walter Boughton,et al.  The Australian water balance model , 2004, Environ. Model. Softw..

[12]  A. Young,et al.  Stream flow simulation within UK ungauged catchments using a daily rainfall-runoff model , 2006 .

[13]  D. Posada,et al.  Model selection and model averaging in phylogenetics: advantages of akaike information criterion and bayesian approaches over likelihood ratio tests. , 2004, Systematic biology.

[14]  Venkat Lakshmi,et al.  Predictions in ungauged basins as a catalyst for multidisciplinary hydrology , 2004 .

[15]  R. Darlington,et al.  Regression and Linear Models , 1990 .

[16]  Martyn P. Clark,et al.  Framework for Understanding Structural Errors (FUSE): A modular framework to diagnose differences between hydrological models , 2008 .

[17]  Simon G Thompson,et al.  Structural and parameter uncertainty in Bayesian cost-effectiveness models , 2010, Journal of the Royal Statistical Society. Series C, Applied statistics.

[18]  B. Bates,et al.  A Markov Chain Monte Carlo Scheme for parameter estimation and inference in conceptual rainfall‐runoff modeling , 2001 .

[19]  Murugesu Sivapalan,et al.  Scale issues in hydrological modelling: A review , 1995 .

[20]  Jeroen P. van der Sluijs,et al.  A framework for dealing with uncertainty due to model structure error , 2004 .

[21]  Montserrat Fuentes,et al.  Model Evaluation and Spatial Interpolation by Bayesian Combination of Observations with Outputs from Numerical Models , 2005, Biometrics.

[22]  D. Kavetski,et al.  Towards a Bayesian total error analysis of conceptual rainfall-runoff models: Characterising model error using storm-dependent parameters , 2006 .

[23]  Lucy Marshall,et al.  Towards dynamic catchment modelling: a Bayesian hierarchical mixtures of experts framework , 2007 .

[24]  S. Sorooshian,et al.  Effective and efficient algorithm for multiobjective optimization of hydrologic models , 2003 .

[25]  George Kuczera,et al.  Critical evaluation of parameter consistency and predictive uncertainty in hydrological modeling: A case study using Bayesian total error analysis , 2009 .

[26]  George Kuczera,et al.  Monte Carlo assessment of parameter uncertainty in conceptual catchment models: the Metropolis algorithm , 1998 .

[27]  Ashish Sharma,et al.  Modeling the catchment via mixtures: Issues of model specification and validation , 2005 .

[28]  Henrik Madsen,et al.  An evaluation of the impact of model structure on hydrological modelling uncertainty for streamflow simulation , 2004 .

[29]  Patrick M. Reed,et al.  Reducing uncertainty in predictions in ungauged basins by combining hydrologic indices regionalization and multiobjective optimization , 2008, Water Resources Research.

[30]  R. Moore.  The PDM rainfall-runoff model , 2007 .

[31]  Robert A. Jacobs,et al.  A Bayesian Approach to Model Selection in Hierarchical Mixtures-of-Experts Architectures , 1997, Neural Networks.

[32]  T. McMahon,et al.  Application of the daily rainfall-runoff model MODHYDROLOG to 28 Australian catchments , 1994 .

[33]  C C Drovandi,et al.  Estimation of Parameters for Macroparasite Population Evolution Using Approximate Bayesian Computation , 2011, Biometrics.

[34]  Mark M. Tanaka,et al.  Sequential Monte Carlo without likelihoods , 2007, Proceedings of the National Academy of Sciences.

[35]  Ashish Sharma,et al.  A comparative study of Markov chain Monte Carlo methods for conceptual rainfall‐runoff modeling , 2004 .

[36]  Yuqiong Liu,et al.  Uncertainty in hydrologic modeling: Toward an integrated data assimilation framework , 2007 .

[37]  Hoshin Vijai Gupta,et al.  Regionalization of constraints on expected watershed response behavior for improved predictions in ungauged basins , 2007 .

[38]  Michael I. Jordan,et al.  Hierarchical Mixtures of Experts and the EM Algorithm , 1994, Neural Computation.

[39]  Keith Beven,et al.  The future of distributed models: model calibration and uncertainty prediction. , 1992 .

[40]  Fengchun Peng,et al.  Bayesian Inference in Mixtures-of-Experts and Hierarchical Mixtures-of-Experts Models With an Applic , 1996 .

[41]  David Draper,et al.  Assessment and Propagation of Model Uncertainty , 2011 .

[42]  Luis R. Pericchi,et al.  A case for a reassessment of the risks of extreme hydrological hazards in the Caribbean , 2006 .

[43]  M. P. Wand,et al.  Generalised linear mixed model analysis via sequential Monte Carlo sampling , 2008, 0810.1163.

[44]  Dmitri Kavetski,et al.  Elements of a flexible approach for conceptual hydrological modeling: 1. Motivation and theoretical development , 2011 .

[45]  Tyler Smith,et al.  Bayesian methods in hydrologic modeling: A study of recent advancements in Markov chain Monte Carlo techniques , 2008 .

[46]  Scott A. Sisson,et al.  Efficient hydrological model parameter optimization with Sequential Monte Carlo sampling , 2012, Environ. Model. Softw..

[47]  Hoshin Vijai Gupta,et al.  Model identification for hydrological forecasting under uncertainty , 2005 .

[48]  J. McDonnell,et al.  Constraining dynamic TOPMODEL responses for imprecise water table information using fuzzy rule based performance measures , 2004 .

[49]  Dmitri Kavetski,et al.  Elements of a flexible approach for conceptual hydrological modeling: 2. Application and experimental insights , 2011 .

[50]  Robert Tibshirani,et al.  The Elements of Statistical Learning: Data Mining, Inference, and Prediction, 2nd Edition , 2001, Springer Series in Statistics.

[51]  P. Moral,et al.  Sequential Monte Carlo samplers , 2002, cond-mat/0212648.

[52]  Kuolin Hsu,et al.  Uncertainty assessment of hydrologic model states and parameters: Sequential data assimilation using the particle filter , 2005 .

[53]  Neil J. Gordon,et al.  A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking , 2002, IEEE Trans. Signal Process..

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

[55]  George Kuczera,et al.  Bayesian analysis of input uncertainty in hydrological modeling: 2. Application , 2006 .

[56]  Ashish Sharma,et al.  Bayesian calibration and uncertainty analysis of hydrological models: A comparison of adaptive Metropolis and sequential Monte Carlo samplers , 2011 .

[57]  Adrian E. Raftery,et al.  Bayesian model averaging: a tutorial (with comments by M. Clyde, David Draper and E. I. George, and a rejoinder by the authors , 1999 .

[58]  K. Beven,et al.  Model Calibration and Uncertainty Estimation , 2006 .

[59]  P. E. O'connell,et al.  IAHS Decade on Predictions in Ungauged Basins (PUB), 2003–2012: Shaping an exciting future for the hydrological sciences , 2003 .

[60]  Soroosh Sorooshian,et al.  Toward improved calibration of hydrologic models: Multiple and noncommensurable measures of information , 1998 .

[61]  A. Shamseldin,et al.  Methods for combining the outputs of different rainfall–runoff models , 1997 .