Evaluating post-processing approaches for monthly and seasonal streamflow forecasts

Abstract. Streamflow forecasting is prone to substantial uncertainty due to errors in meteorological forecasts, hydrological model structure, and parameterization, as well as in the observed rainfall and streamflow data used to calibrate the models. Statistical streamflow post-processing is an important technique available to improve the probabilistic properties of the forecasts. This study evaluates post-processing approaches based on three transformations – logarithmic (Log), log-sinh (Log-Sinh), and Box–Cox with λ=0.2 (BC0.2) – and identifies the best-performing scheme for post-processing monthly and seasonal (3-months-ahead) streamflow forecasts, such as those produced by the Australian Bureau of Meteorology. Using the Bureau's operational dynamic streamflow forecasting system, we carry out comprehensive analysis of the three post-processing schemes across 300 Australian catchments with a wide range of hydro-climatic conditions. Forecast verification is assessed using reliability and sharpness metrics, as well as the Continuous Ranked Probability Skill Score (CRPSS). Results show that the uncorrected forecasts (i.e. without post-processing) are unreliable at half of the catchments. Post-processing of forecasts substantially improves reliability, with more than 90 % of forecasts classified as reliable. In terms of sharpness, the BC0.2 scheme substantially outperforms the Log and Log-Sinh schemes. Overall, the BC0.2 scheme achieves reliable and sharper-than-climatology forecasts at a larger number of catchments than the Log and Log-Sinh schemes. The improvements in forecast reliability and sharpness achieved using the BC0.2 post-processing scheme will help water managers and users of the forecasting service make better-informed decisions in planning and management of water resources. Highlights. Uncorrected and post-processed streamflow forecasts (using three transformations, namely Log, Log-Sinh, and BC0.2) are evaluated over 300 diverse Australian catchments. Post-processing enhances streamflow forecast reliability, increasing the percentage of catchments with reliable predictions from 50 % to over 90 %. The BC0.2 transformation achieves substantially better forecast sharpness than the Log-Sinh and Log transformations, particularly in dry catchments.

[1]  C. M. DeChant,et al.  Improving the characterization of initial condition for ensemble streamflow prediction using data assimilation , 2011 .

[2]  Rebecca E. Morss,et al.  Examining the use of weather forecasts in decision scenarios: results from a US survey with implications for uncertainty communication , 2010 .

[3]  D. Cox,et al.  An Analysis of Transformations , 1964 .

[4]  Quan J. Wang,et al.  Post-processing rainfall forecasts from numerical weather prediction models for short-term streamflow forecasting , 2013 .

[5]  C. Perrin,et al.  Improvement of a parsimonious model for streamflow simulation , 2003 .

[6]  Tyler Smith,et al.  Modeling residual hydrologic errors with Bayesian inference , 2015 .

[7]  James C. Bennett,et al.  Reliable long‐range ensemble streamflow forecasts: Combining calibrated climate forecasts with a conceptual runoff model and a staged error model , 2016 .

[8]  K. Georgakakos,et al.  Assessment of Folsom lake response to historical and potential future climate scenarios: 1. Forecasting , 2001 .

[9]  A. Raftery,et al.  Probabilistic forecasts, calibration and sharpness , 2007 .

[10]  Anton H. Westveld,et al.  Calibrated Probabilistic Forecasting Using Ensemble Model Output Statistics and Minimum CRPS Estimation , 2005 .

[11]  A. H. Murphy,et al.  On the Relationship between the Accuracy and Value of Forecasts in the Cost–Loss Ratio Situation , 1987 .

[12]  James C. Bennett,et al.  Assessment of an ensemble seasonal streamflow forecasting system for Australia , 2017 .

[13]  Dominic A. Hudson,et al.  Improving Intraseasonal Prediction with a New Ensemble Generation Strategy , 2013 .

[14]  Q. J. Wang,et al.  Multisite probabilistic forecasting of seasonal flows for streams with zero value occurrences , 2011 .

[15]  Yonghong Yin,et al.  An Ensemble Ocean Data Assimilation System for Seasonal Prediction , 2011 .

[16]  Chong-Yu Xu,et al.  Observed changes of drought/wetness episodes in the Pearl River basin, China, using the standardized precipitation index and aridity index , 2009 .

[17]  S. Galelli,et al.  Complex relationship between seasonal streamflow forecast skill and value in reservoir operations , 2017 .

[18]  Quan J. Wang,et al.  Improving statistical forecasts of seasonal streamflows using hydrological model output , 2012 .

[19]  Hamid Moradkhani,et al.  Towards improved post‐processing of hydrologic forecast ensembles , 2014 .

[20]  Holger R. Maier,et al.  A hybrid approach to monthly streamflow forecasting: Integrating hydrological model outputs into a Bayesian artificial neural network , 2016 .

[21]  Quan J. Wang,et al.  Overview of Communication Strategies for Uncertainty in Hydrological Forecasting in Australia , 2019, Handbook of Hydrometeorological Ensemble Forecasting.

[22]  Q. Shao,et al.  Experimental evaluation of the dynamic seasonal streamflow forecasting approach , 2011 .

[23]  John O. Carter,et al.  Using spatial interpolation to construct a comprehensive archive of Australian climate data , 2001, Environ. Model. Softw..

[24]  Vincent Boucher,et al.  Moving beyond the cost–loss ratio: economic assessment of streamflow forecasts for a risk-averse decision maker , 2016 .

[25]  George Kuczera,et al.  Comparison of joint versus postprocessor approaches for hydrological uncertainty estimation accounting for error autocorrelation and heteroscedasticity , 2014 .

[26]  George Kuczera,et al.  Toward a reliable decomposition of predictive uncertainty in hydrological modeling: Characterizing rainfall errors using conditional simulation , 2011 .

[27]  D. Lettenmaier,et al.  Use of satellite snow-cover data for streamflow prediction in the Feather River Basin, California , 2010 .

[28]  George Kuczera,et al.  Improving probabilistic prediction of daily streamflow by identifying Pareto optimal approaches for modeling heteroscedastic residual errors , 2017 .

[29]  Dynamical seasonal prediction of climate extremes in the Pacific , 2013 .

[30]  Anthony J. Jakeman,et al.  Performance of conceptual rainfall‐runoff models in low‐yielding ephemeral catchments , 1997 .

[31]  Thomas Nipen,et al.  Reliable probabilistic forecasts from an ensemble reservoir inflow forecasting system , 2014 .

[32]  Wade T. Crow,et al.  Improving Streamflow Prediction Using Remotely-Sensed Soil Moisture and Snow Depth , 2016, Remote. Sens..

[33]  M. Sivapalan,et al.  Characterizing hydrologic change through catchment classification , 2014 .

[34]  Quan J. Wang,et al.  A Bayesian joint probability modeling approach for seasonal forecasting of streamflows at multiple sites , 2009 .

[35]  Peter Krause,et al.  Combined uncertainty of hydrological model complexity and satellite-based forcing data evaluated in two data-scarce semi-arid catchments in Ethiopia , 2014 .

[36]  M. England,et al.  An analysis of late twentieth century trends in Australian rainfall , 2009 .

[37]  H. Hersbach.  Decomposition of the Continuous Ranked Probability Score for Ensemble Prediction Systems , 2000 .

[38]  Tempei Hashino,et al.  Evaluation of bias-correction methods for ensemble streamflow volume forecasts , 2006 .

[39]  James D. Brown,et al.  Verification of temperature, precipitation, and streamflow forecasts from the NOAA/NWS Hydrologic Ensemble Forecast Service (HEFS): 1. Experimental design and forcing verification , 2014 .

[40]  F. Pappenberger,et al.  Bias correcting precipitation forecasts to improve the skill of seasonal streamflow forecasts , 2016 .

[41]  A. P. Dawid,et al.  Present position and potential developments: some personal views , 1984 .

[42]  Dmitri Kavetski,et al.  State updating and calibration period selection to improve dynamic monthly streamflow forecasts for an environmental flow management application , 2018 .

[43]  Dmitri Kavetski,et al.  Pursuing the method of multiple working hypotheses for hydrological modeling , 2011 .

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

[45]  C. Tyralla,et al.  Incorporating structural uncertainty of hydrological models in likelihood functions via an ensemble range approach , 2016 .

[46]  F. Pappenberger,et al.  Operational European Flood Forecasting , 2013 .

[47]  B. Mcavaney,et al.  An analogue-based method to downscale surface air temperature: application for Australia , 2001 .

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

[49]  J. Schaake,et al.  Correcting Errors in Streamflow Forecast Ensemble Mean and Spread , 2008 .

[50]  Martin L. Hazelton,et al.  Methods of Moments Estimation , 2011, International Encyclopedia of Statistical Science.

[51]  M. Clark,et al.  An intercomparison of approaches for improving operational seasonal streamflow forecasts , 2017 .

[52]  R. Weingartner,et al.  Monthly streamflow forecasting at varying spatial scales in the Rhine basin , 2017 .

[53]  K. Bogner,et al.  Error‐correction methods and evaluation of an ensemble based hydrological forecasting system for the Upper Danube catchment , 2008 .

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

[55]  Quan J. Wang,et al.  A Bayesian joint probability post-processor for reducing errors and quantifying uncertainty in monthly streamflow predictions , 2012 .

[56]  Quan J. Wang,et al.  A log‐sinh transformation for data normalization and variance stabilization , 2012 .

[57]  Harold E. Brooks,et al.  On the relationship between Clayton's skill score and expected value for forecasts of binary events , 2002 .

[58]  M. Sivapalan,et al.  Catchment classification: hydrological analysis of catchment behavior through process-based modeling along a climate gradient , 2011 .

[59]  Quan J. Wang,et al.  Selecting predictors for seasonal streamflow predictions using a Bayesian joint probability (BJP) modelling approach , 2009 .

[60]  S. Shapiro,et al.  An Analysis of Variance Test for Normality (Complete Samples) , 1965 .

[61]  George Kuczera,et al.  Bayesian analysis of input uncertainty in hydrological modeling: 1. Theory , 2006 .

[62]  Quan J. Wang,et al.  A Bayesian Approach to Predictor Selection for Seasonal Streamflow Forecasting , 2012 .

[63]  Quan J. Wang,et al.  Ensemble forecasting of sub-seasonal to seasonal streamflow by a Bayesian joint probability modelling approach , 2016 .

[64]  Andreas Scheidegger,et al.  Improving uncertainty estimation in urban hydrological modeling by statistically describing bias , 2013 .

[65]  W. Briggs.  Statistical Methods in the Atmospheric Sciences , 2007 .

[66]  V. Bell,et al.  Hydrological Outlook UK: an operational streamflow and groundwater level forecasting system at monthly to seasonal time scales , 2017 .

[67]  Seasonal Streamflow Forecasting with a workflow-based dynamic hydrologic modelling approach , 2011 .

[68]  Stefania Tamea,et al.  Verification tools for probabilistic forecasts of continuous hydrological variables , 2006 .

[69]  James D. Brown,et al.  The Science of NOAA's Operational Hydrologic Ensemble Forecast Service , 2014 .

[70]  James C. Bennett,et al.  Error reduction and representation in stages (ERRIS) in hydrologicalmodelling for ensemble streamflow forecasting , 2016 .