Ensemble flood forecasting: A review
01 May 2010-pp 15247
TL;DR: The scientific drivers of this shift towards ‘ensemble flood forecasting’ and the literature evidence of the ‘added value’ of flood forecasts based on EPS are reviewed.
Abstract: Operational medium range flood forecasting systems are increasingly moving towards the adoption of ensembles of numerical weather predictions (NWP), known as ensemble prediction systems (EPS), to drive their predictions. We review the scientific drivers of this shift towards such ‘ensemble flood forecasting’ and discuss several of the questions surrounding best practice in using EPS in flood forecasting systems. We also review the literature evidence of the ‘added value’ of flood forecasts based on EPS and point to remaining key challenges in using EPS successfully.
Citations
More filters
01 Dec 2010
TL;DR: In this article, the authors modeled the dynamic evolution of uncertainties involved in the various forecast products and explored their effect on real-time reservoir operation decisions, showing that forecast uncertainty exerts significant effects.
Abstract: Various hydrological forecast products have been applied to real-time reservoir operation, including deterministic streamflow forecast (DSF), DSF-based probabilistic streamflow forecast (pseudo-PSF, pPSF), and ensemble or probabilistic streamflow forecast (denoted as real-PSF, rPSF). DSF represents forecast uncertainty in the form of deterministic forecast errors, pPSF a conditional distribution of forecast uncertainty for a given DSF, and rPSF a probabilistic uncertainty distribution. Compared to previous studies that treat the forecast products as ad hoc inputs for reservoir operation models, this paper attempts to model the dynamic evolution of uncertainties involved in the various forecast products and explores their effect on real-time reservoir operation decisions. Through a hypothetical example of a single-objective real-time reservoir operation model, the results illustrate that forecast uncertainty exerts significant effects. Reservoir operation efficiency, as measured by a utility function, decreases as the forecast uncertainty increases but the magnitude depends on the forecast products used. In general, the utility of the reservoir operation with rPSF is nearly as high as the utility obtained with a perfect forecast. Meanwhile, the utilities of DSF and pPSF are similar to each other but not as high as rPSF. Moreover, streamflow variability and reservoir capacity can change the magnitude of the effects of forecast uncertainty, but not the relative merit of DSF, pPSF, and rPSF.
132 citations
Journal Article•
TL;DR: To truly assess model performance, it is important to compare one’s results with results obtained in some other way, i.e. to choose an appropriate benchmark series.
Abstract: Hydrological models are applied frequently to scientific or practical problems. For many applications it is concluded that the model has been able to reproduce the measurements with ‘acceptable accuracy’. The question is what we mean by this term; the meaning of ‘acceptable accuracy’ can be quite subjective. We might compute statistical goodness-of-fit measures such as model efficiency, but even the use of such a measure does not necessarily allow an objective judgment of model performance. Does an efficiency of 0.8 for the runoff simulations indicate good or poor model performance? The answer depends on whom you ask. But it also depends on what could be achieved given the specific catchment and the observed data. What might be a poor fit for a watershed with excellent measurements might rightly be considered to be good in a watershed where the available data are of poor quality. To truly assess model performance, it is important to compare one’s results with results obtained in some other way, i.e. to choose an appropriate benchmark series.
117 citations
01 Apr 2016
TL;DR: In this paper, a pair of hydrometeorological modeling systems were calibrated and evaluated for the Ayalon basin in central Israel to assess the advantages and limitations of one-way versus two-way coupled modeling systems for flood prediction.
Abstract: A pair of hydro-meteorological modeling systems were calibrated and evaluated for the Ayalon basin in central Israel to assess the advantages and limitations of one-way versus two-way coupled modeling systems for flood prediction. The models used included the Hydrological Engineering Center-Hydrological Modeling System (HEC-HMS) model and the Weather Research and Forecasting (WRF) Hydro modeling system. The models were forced by observed, interpolated precipitation from rain-gauges within the basin, and with modeled precipitation from the WRF atmospheric model. Detailed calibration and evaluation was carried out for two major winter storms in January and December 2013. Then, both modeling systems were executed and evaluated in an operational mode for the full 2014/2015 rainy season. Outputs from these simulations were compared to observed measurements from the hydrometric station at the Ayalon basin outlet. Various statistical metrics were employed to quantify and analyze the results: correlation, Root Mean Square Error (RMSE) and the Nash–Sutcliffe (NS) efficiency coefficient. Foremost, the results presented in this study highlight the sensitivity of hydrological responses to different sources of simulated and observed precipitation data, and demonstrate improvement, although not significant, at the Hydrological response, like simulated hydrographs. With observed precipitation data both calibrated models closely simulated the observed hydrographs. The two-way coupled WRF/WRF-Hydro modeling system produced improved both the precipitation and hydrological simulations as compared to the one-way WRF simulations. Findings from this study, as well as previous studies, suggest that the use of two-way atmospheric-hydrological coupling has the potential to improve precipitation and, therefore, hydrological forecasts for early flood warning applications. However, more research needed in order to better understand the land-atmosphere coupling mechanisms driving hydrometeorological processes on a wider variety precipitation and terrestrial hydrologic systems.
61 citations
Cites background from "Ensemble flood forecasting: A revie..."
...2008 [21]; Cloke and Pappenberger, 2009 [22]; Sin Shiha et al....
[...]
Journal Article•
TL;DR: Evaluating a Bayesian Forecasting System for real-time flood forecasting reveals the crucial role of the hydrological uncertainty processor when compensating, at the cost of wider forecast intervals, for the unreliable and biased predictive distribution resulting from the Precipitation Uncertainty Processor.
Abstract: Summary The paper evaluates, for a number of flood events, the performance of a Bayesian Forecasting System (BFS), with the aim of evaluating total uncertainty in real-time flood forecasting. The predictive uncertainty of future streamflow is estimated through the Bayesian integration of two separate processors. The former evaluates the propagation of input uncertainty on simulated river discharge, the latter computes the hydrological uncertainty of actual river discharge associated with all other possible sources of error. A stochastic model and a distributed rainfall–runoff model were assumed, respectively, for rainfall and hydrological response simulations. A case study was carried out for a small basin in the Calabria region (southern Italy). The performance assessment of the BFS was performed with adequate verification tools suited for probabilistic forecasts of continuous variables such as streamflow. Graphical tools and scalar metrics were used to evaluate several attributes of the forecast quality of the entire time-varying predictive distributions: calibration, sharpness, accuracy, and continuous ranked probability score (CRPS). Besides the overall system, which incorporates both sources of uncertainty, other hypotheses resulting from the BFS properties were examined, corresponding to (i) a perfect hydrological model; (ii) a non-informative rainfall forecast for predicting streamflow; and (iii) a perfect input forecast. The results emphasize the importance of using different diagnostic approaches to perform comprehensive analyses of predictive distributions, to arrive at a multifaceted view of the attributes of the prediction. For the case study, the selected criteria revealed the interaction of the different sources of error, in particular the crucial role of the hydrological uncertainty processor when compensating, at the cost of wider forecast intervals, for the unreliable and biased predictive distribution resulting from the Precipitation Uncertainty Processor.
45 citations
01 Dec 2013
TL;DR: Wang et al. as discussed by the authors analyzed the reservoir real-time operating risk with ensemble-based hydrologic forecasts as inputs and developed a method by using the forecast horizon point to divide the future time into two stages, the forecast lead-time and the unpredicted time.
Abstract: Quantitative analysis of the risk for reservoir real-time operation is a hard task owing to the difficulty of accurate description of inflow uncertainties. The ensemble-based hydrologic forecasts directly depict the inflows not only the marginal distributions but also their persistence via scenarios. This motivates us to analyze the reservoir real-time operating risk with ensemble-based hydrologic forecasts as inputs. A method is developed by using the forecast horizon point to divide the future time into two stages, the forecast lead-time and the unpredicted time. The risk within the forecast lead-time is computed based on counting the failure number of forecast scenarios, and the risk in the unpredicted time is estimated using the reservoir routing with the design flood hydrographs. As a result, a two-stage risk analysis method is set up to quantify the entire flood risks by defining the ratio of the number of failure scenarios (excessive the critical value) to the total scenarios number. The China’s Three Gorges Reservoir is selected as a case study, where the parameter and precipitation uncertainties are conducted to produce ensemble-based hydrologic forecasts. Two reservoir operation schemes, the historical operated and scenario optimization, are evaluated with the flood risks and hydropower profits analysis. The derived risk, which units with yearly scale, associates with the flood protection standards (described with return periods) that can be used as the acceptable level to operate reservoir. With the 2010 flood, it is found that the proposed method can greatly improve the hydropower generation with acceptable flood risks.
40 citations
Cites methods from "Ensemble flood forecasting: A revie..."
...The ensemble streamflow prediction is a general and popular forecast technique for the real reservoir operation (Alemu et al. 2011), which has been well literature by Cloke and Pappenberger (2009)....
[...]
References
More filters
TL;DR: The GLUE procedure works with multiple sets of parameter values and allows that, within the limitations of a given model structure and errors in boundary conditions and field observations, different sets of values may be equally likely as simulators of a catchment.
Abstract: This paper describes a methodology for calibration and uncertainty estimation of distributed models based on generalized likelihood measures. The GLUE procedure works with multiple sets of parameter values and allows that, within the limitations of a given model structure and errors in boundary conditions and field observations, different sets of values may be equally likely as simulators of a catchment. Procedures for incorporating different types of observations into the calibration; Bayesian updating of likelihood values and evaluating the value of additional observations to the calibration process are described. The procedure is computationally intensive but has been implemented on a local parallel processing computer.
4,146 citations
"Ensemble flood forecasting: A revie..." refers methods in this paper
...Alternatives to the formal treatment of cascading uncertainties includes a generalized Bayesian approach based on the GLUE methodology (Beven and Binley, 1992) presented by Pappenberger et al....
[...]
16 Dec 2011
TL;DR: Jolliffe et al. as mentioned in this paper proposed a framework for verification of spatial fields based on binary and categorical events, and proved the correctness of the proposed framework with past, present and future predictions.
Abstract: List of Contributors. Preface. 1. Introduction (I. Jolliffe & D. Stephenson). 2. Basic Concepts (J. Potts). 3. Binary Events (I. Mason). 4. Categorical Events (R. Livezey). 5. Continuous Variables (M. Deque). 6. Verification of Spatial Fields (W. Drosdowsky & H. Zhang). 7. Probability and Ensemble Forecasts (Z. Toth, et al.). 8. Economic Value and Skill (D. Richardson). 9. Forecast Verification: Past, Present and Future (D. Stephenson & I. Jolliffe). Glossary. References. Author Index. Subject Index.
1,633 citations
TL;DR: The Ensemble Transform Kalman Filter (ET KF) as discussed by the authors is a suboptimal Kalman filter that uses ensemble transformation and a normalization to obtain the prediction error covariance matrix associated with a particular deployment of observational resources.
Abstract: A suboptimal Kalman filter called the ensemble transform Kalman filter (ET KF) is introduced. Like other Kalman filters, it provides a framework for assimilating observations and also for estimating the effect of observations on forecast error covariance. It differs from other ensemble Kalman filters in that it uses ensemble transformation and a normalization to rapidly obtain the prediction error covariance matrix associated with a particular deployment of observational resources. This rapidity enables it to quickly assess the ability of a large number of future feasible sequences of observational networks to reduce forecast error variance. The ET KF was used by the National Centers for Environmental Prediction in the Winter Storm Reconnaissance missions of 1999 and 2000 to determine where aircraft should deploy dropwindsondes in order to improve 24‐72-h forecasts over the continental United States. The ET KF may be applied to any well-constructed set of ensemble perturbations. The ET KF technique supercedes the ensemble transform (ET) targeting technique of Bishop and Toth. In the ET targeting formulation, the means by which observations reduced forecast error variance was not expressed mathematically. The mathematical representation of this process provided by the ET KF enables such things as the evaluation of the reduction in forecast error variance associated with individual flight tracks and assessments of the value of targeted observations that are distributed over significant time intervals. It also enables a serial targeting methodology whereby one can identify optimal observing sites given the location and error statistics of other observations. This allows the network designer to nonredundantly position targeted observations. Serial targeting can also be used to greatly reduce the computations required to identify optimal target sites. For these theoretical and practical reasons, the ET KF technique is more useful than the ET technique. The methodology is illustrated with observation system simulation experiments involving a barotropic numerical model of tropical cyclonelike vortices. These include preliminary empirical tests of ET KF predictions using ET KF, 3DVAR, and hybrid data assimilation schemes—the results of which look promising. To concisely describe the future feasible sequences of observations considered in adaptive sampling problems, an extension to Ide et al.’s unified notation for data assimilation is suggested.
1,338 citations
TL;DR: In this article, the continuous ranked probability score (CRPS) is decomposed into a reliability part and a resolution/uncertainty part, in a way similar to the decomposition of the Brier score.
Abstract: Some time ago, the continuous ranked probability score (CRPS) was proposed as a new verification tool for (probabilistic) forecast systems. Its focus is on the entire permissible range of a certain (weather) parameter. The CRPS can be seen as a ranked probability score with an infinite number of classes, each of zero width. Alternatively, it can be interpreted as the integral of the Brier score over all possible threshold values for the parameter under consideration. For a deterministic forecast system the CRPS reduces to the mean absolute error. In this paper it is shown that for an ensemble prediction system the CRPS can be decomposed into a reliability part and a resolution/uncertainty part, in a way that is similar to the decomposition of the Brier score. The reliability part of the CRPS is closely connected to the rank histogram of the ensemble, while the resolution/ uncertainty part can be related to the average spread within the ensemble and the behavior of its outliers. The usefulness of such a decomposition is illustrated for the ensemble prediction system running at the European Centre for Medium-Range Weather Forecasts. The evaluation of the CRPS and its decomposition proposed in this paper can be extended to systems issuing continuous probability forecasts, by realizing that these can be interpreted as the limit of ensemble forecasts with an infinite number of members.
1,148 citations
TL;DR: In this paper, a stochastic representation of random error associated with parametrized physical processes is described, and its impact in the European Centre for Medium-Range Weather Forecasts Ensemble Prediction System (ECMWF EPS) is discussed.
Abstract: SUMMARY A stochastic representation of random error associated with parametrized physical processes (‘stochastic physics’) is described, and its impact in the European Centre for Medium-Range Weather Forecasts Ensemble Prediction System (ECMWF EPS) is discussed. Model random errors associated with physical parametrizations are simulated by multiplying the total parametrized tendencies by a random number sampled from a uniform distribution between 0.5 and 1.5. A number of diagnostics are described and a choice of parameters is made. It is shown how the scheme increases the spread of the ensemble, and improves the skill of the probabilistic prediction of weather parameters such as precipitation. A choice of stochastic parameters is made for operational implementation. The scheme was implemented successfully in the operational ECMWF EPS on 21 October 1998.
1,067 citations
"Ensemble flood forecasting: A revie..." refers background in this paper
...Some EPS also additionally incorporate parameter uncertainty in the generation of the ensemble forecasts (Buizza et al., 1999; Houtekamer and Lefaivre, 1997; Shutts, 2005)....
[...]
...Weather forecasts remain limited by not only the numerical representation of the physical processes, but also the resolution of the simulated atmospheric dynamics and the sensitivity of the solutions to the pattern of initial conditions and sub-grid parameterization (Buizza et al., 1999)....
[...]
...It is often thought that a substantial increase in resolution of the models will allow resolution of rainfall cells in predictions and thus remove some of the large errors (Buizza et al., 1999; Undén, 2006)....
[...]