Generalized autoregressive conditional heteroscedasticity modelling of hydrologic time series

The existence of time‐dependent variance or conditional variance, commonly called heteroscedasticity, in hydrologic time series has not been thoroughly investigated. This paper deals with modelling the heteroscedasticity in the residuals of the seasonal autoregressive integrated moving average (SARIMA) model using a generalized autoregressive conditional heteroscedasticity (GARCH) model. The model is applied to two monthly rainfall time series from humid and arid regions. The effect of Box–Cox transformation and seasonal differencing on the remaining seasonal heteroscedasticity in the residuals of the SARIMA model is also investigated. It is shown that the seasonal heteroscedasticity in the residuals of the SARIMA model can be removed using Box–Cox transformation along with seasonal differencing for the humid region rainfall. On the other hand, transformation and seasonal differencing could not remove heteroscedasticity from the residuals of the SARIMA model fitted to rainfall data in the arid region. Therefore, the GARCH modelling approach is necessary to capture the heteroscedasticity remaining in the residuals of a SARIMA model. However, the evaluation criteria do not necessarily show that the GARCH model improves the performance of the SARIMA model. Copyright © 2012 John Wiley & Sons, Ltd.     

Impact of the numbers of observations and calibration parameters on equifinality,model performance,and output and parameter uncertainty

The level of model complexity that can be effectively supported by available information has long been a subject of many studies in hydrologic modelling. In particular, distributed parameter models tend to be regarded as overparameterized because of numerous parameters used to describe spatially heterogeneous hydrologic processes. However, it is not clear how parameters and observations influence the degree of overparameterization, equifinality of parameter values, and uncertainty. This study investigated the impact of the numbers of observations and parameters on calibration quality including equifinality among calibrated parameter values, model performance, and output/parameter uncertainty using the Soil and Water Assessment Tool model. In the experiments, the number of observations was increased by expanding the calibration period or by including measurements made at inner points of a watershed. Similarly, additional calibration parameters were included in the order of their sensitivity. Then, unique sets of parameters were calibrated with the same objective function, optimization algorithm, and stopping criteria but different numbers of observations. The calibration quality was quantified with statistics calculated based on the ‘behavioural’ parameter sets, identified using 1% and 5% cut‐off thresholds in a generalized likelihood uncertainty estimation framework. The study demonstrated that equifinality, model performance, and output/parameter uncertainty were responsive to the numbers of observations and calibration parameters; however, the relationship between the numbers, equifinality, and uncertainty was not always conclusive. Model performance improved with increased numbers of calibration parameters and observations, and substantial equifinality did neither necessarily mean bad model performance nor large uncertainty in the model outputs and parameters. Copyright © 2015 John Wiley & Sons, Ltd.     

Model uncertainty analysis by variance decomposition

Errors and uncertainties in hydrological, hydraulic and environmental models are often substantial. In good modelling practice, they are quantified in order to supply decision-makers with important additional information on model limitations and sources of uncertainty. Several uncertainty analysis methods exist, often with various underlying assumptions. One of these methods is based on variance decomposition. The method allows splitting the variance of the total error in the model results (as estimated after comparing model results with observations) in its major contributing uncertainty sources. This paper discusses an advanced version of that method where error distributions for rainfall, other inputs and parameters are propagated in the model and the “rest” uncertainties considered as model structural errors for different parts of the model. By expert knowledge, the iid assumption that is often made in model error analysis is addressed upfront. The method also addresses the problems of heteroscedasticity and serial dependence of the errors involved. The method has been applied by the author to modelling applications of sewer water quantity and quality, river water quality and river flooding.     

Probabilistic calibration of a distributed hydrological model for flood forecasting

Abstract

The complexity of distributed hydrological models has led to improvements in calibration methodologies in recent years. There are various manual, automatic and hybrid methods of calibration. Most use a single objective function to calculate estimation errors. The use of multi-objective calibration improves results, since different aspects of the hydrograph may be considered simultaneously. However, the uncertainty of estimates from a hydrological model can only be taken into account by using a probabilistic approach. This paper presents a calibration method of probabilistic nature, based on the determination of probability functions that best characterize different parameters of the model. The method was applied to the Real-time Interactive Basin Simulator (RIBS) distributed hydrological model using the Manzanares River basin in Spain as a case study. The proposed method allows us to consider the uncertainty in the model estimates by obtaining the probability distributions of flows in the flood hydrograph.

Citation Mediero, L., Garrote, L. & Martín-Carrasco, F. J. (2011) Probabilistic calibration of a distributed hydrological model for flood forecasting. Hydrol. Sci. J. 56(7), 1129–1149.     

Hydrologic data assimilation using particle Markov chain Monte Carlo simulation: Theory,concepts and applications

During the past decades much progress has been made in the development of computer based methods for parameter and predictive uncertainty estimation of hydrologic models. The goal of this paper is twofold. As part of this special anniversary issue we first shortly review the most important historical developments in hydrologic model calibration and uncertainty analysis that has led to current perspectives. Then, we introduce theory, concepts and simulation results of a novel data assimilation scheme for joint inference of model parameters and state variables. This Particle-DREAM method combines the strengths of sequential Monte Carlo sampling and Markov chain Monte Carlo simulation and is especially designed for treatment of forcing, parameter, model structural and calibration data error. Two different variants of Particle-DREAM are presented to satisfy assumptions regarding the temporal behavior of the model parameters. Simulation results using a 40-dimensional atmospheric “toy” model, the Lorenz attractor and a rainfall–runoff model show that Particle-DREAM, P-DREAM_(VP) and P-DREAM_(IP) require far fewer particles than current state-of-the-art filters to closely track the evolving target distribution of interest, and provide important insights into the information content of discharge data and non-stationarity of model parameters. Our development follows formal Bayes, yet Particle-DREAM and its variants readily accommodate hydrologic signatures, informal likelihood functions or other (in)sufficient statistics if those better represent the salient features of the calibration data and simulation model used.     

Addressing uncertainty in reflectivity‐rainfall relations in mountain watersheds during summer convection

The use of precipitation estimates from weather radar reflectivity has become widespread in hydrologic predictions. However, uncertainty remains in the use of the nonlinear reflectivity–rainfall (Z‐R) relation, in particular for mountainous regions where ground validation stations are often lacking, land surface data sets are inaccurate and the spatial variability in many features is high. In this study, we assess the propagation of rainfall errors introduced by different Z‐R relations on distributed hydrologic model performance for four mountain basins in the Colorado Front Range. To do so, we compare spatially integrated and distributed rainfall and runoff metrics at seasonal and event time scales during the warm season when convective storms dominate. Results reveal that the basin simulations are quite sensitive to the uncertainties introduced by the Z‐R relation in terms of streamflow, runoff mechanisms and the water balance components. The propagation of rainfall errors into basin responses follows power law relationships that link streamflow uncertainty to the precipitation errors and streamflow magnitude. Overall, different Z‐R relations preserve the spatial distribution of rainfall relative to a reference case, but not the precipitation magnitude, thus leading to large changes in streamflow amounts and runoff spatial patterns at seasonal and event scales. Furthermore, streamflow errors from the Z‐R relation follow a typical pattern that varies with catchment scale where higher uncertainties exist for intermediate‐sized basins. The relatively high error values introduced by two operational Z‐R relations (WSR‐57 and NEXRAD) in terms of the streamflow response indicate that site‐specific Z‐R relations are desirable in the complex terrain region, particularly in light of other uncertainties in the modelling process, such as model parameter values and initial conditions. Copyright © 2012 John Wiley & Sons, Ltd.     

Uncertainty in hydrologic modelling for estimating hydrologic response due to climate change (Santiam River,Oregon)

This paper explores the predicted hydrologic responses associated with the compounded error of cascading global circulation model (GCM) uncertainty through hydrologic model uncertainty due to climate change. A coupled groundwater and surface water flow model (GSFLOW) was used within the differential evolution adaptive metropolis (DREAM) uncertainty approach and combined with eight GCMs to investigate uncertainties in hydrologic predictions for three subbasins of varying hydrogeology within the Santiam River basin in Oregon, USA. Predictions of future hydrology in the Santiam River include increases in runoff in the fall and winter months and decreases in runoff for the spring and summer months. One‐year peak flows were predicted to increase whereas 100‐year peak flows were predicted to slightly decrease. The predicted 10‐year 7‐day low flow decreased in two subbasins with little groundwater influences but increased in another subbasin with substantial groundwater influences. Uncertainty in GCMs represented the majority of uncertainty in the analysis, accounting for an average deviation from the median of 66%. The uncertainty associated with use of GSFLOW produced only an 8% increase in the overall uncertainty of predicted responses compared to GCM uncertainty. This analysis demonstrates the value and limitations of cascading uncertainty from GCM use through uncertainty in the hydrologic model, offers insight into the interpretation and use of uncertainty estimates in water resources analysis, and illustrates the need for a fully nonstationary approach with respect to calibrating hydrologic models and transferring parameters across basins and time for climate change analyses. Copyright © 2012 John Wiley & Sons, Ltd.     

The integrated effects of climate and hydrologic uncertainty on future flood risk assessments

This work examines future flood risk within the context of integrated climate and hydrologic modelling uncertainty. The research questions investigated are (1) whether hydrologic uncertainties are a significant source of uncertainty relative to other sources such as climate variability and change and (2) whether a statistical characterization of uncertainty from a lumped, conceptual hydrologic model is sufficient to account for hydrologic uncertainties in the modelling process. To investigate these questions, an ensemble of climate simulations are propagated through hydrologic models and then through a reservoir simulation model to delimit the range of flood protection under a wide array of climate conditions. Uncertainty in mean climate changes and internal climate variability are framed using a risk‐based methodology and are explored using a stochastic weather generator. To account for hydrologic uncertainty, two hydrologic models are considered, a conceptual, lumped parameter model and a distributed, physically based model. In the conceptual model, parameter and residual error uncertainties are quantified and propagated through the analysis using a Bayesian modelling framework. The approach is demonstrated in a case study for the Coralville Dam on the Iowa River, where recent, intense flooding has raised questions about potential impacts of climate change on flood protection adequacy. Results indicate that the uncertainty surrounding future flood risk from hydrologic modelling and internal climate variability can be of the same order of magnitude as climate change. Furthermore, statistical uncertainty in the conceptual hydrological model can capture the primary structural differences that emerge in flood damage estimates between the two hydrologic models. Copyright © 2014 John Wiley & Sons, Ltd.     

Investigating the uncertainty of satellite altimetry products for hydrodynamic modelling

Satellite altimetry products are increasingly used in many hydraulic applications, and recent studies demonstrate their suitability for the calibration of hydraulic models. The study investigates the effect of satellite‐data uncertainty on the calibration of a quasi‐two‐dimensional (quasi‐2D) model of the middle‐lower portion of the Po river (~140 km). We refer to extended (~16 years of observations) ERS and ENVISAT altimetry products (i.e. River and Lake Hydrology data, RLH) to investigate the effect of (i) record length (i.e. number of satellite measurements at a given satellite track) and (ii) data uncertainty (i.e. altimetry measurements errors) on the calibration of the quasi‐2D model. We first present an assessment of ERS and ENVISAT altimetry errors and then perform the investigations in a Monte Carlo framework by generating datasets of synthetic altimetry products. The results of our analysis further emphasize the suitability of satellite data for the calibration of hydraulic models, providing also a quantitative assessment of the effect of the uncertainty of altimetry products. The analysis highlights the higher accuracy of ENVISAT data, which ensures a stable calibration with ~1.5 years of data (Mean Absolute Error, MAE, lower than 0.4 m, ~0.2 m of which results directly from the uncertainty of ENVISAT data). ERS‐based calibrations become stable with longer series (~3.5–5 years of data), and the negative effect of uncertainty in ERS data is higher (i.e. MAE of 0.6–0.9 m, of which 0.4–0.6 m results from the uncertainty of ERS measurements). Copyright © 2015 John Wiley & Sons, Ltd.     

Distributed and integrated response of a geographic information system‐based hydrologic model in the eastern Palouse region,Idaho

Verification of distributed hydrologic models is rare owing to the lack of spatially detailed field measurements and a common mismatch between the scale at which soil hydraulic properties are measured and the scale of a single modelling unit. In this study, two of the most commonly calibrated parameters, i.e. soil depth and the vertical distribution of lateral saturated hydraulic conductivity K_s, were eliminated by a spatially detailed soil characterization and results of a hillslope‐scale field experiment. The soil moisture routing (SMR) model, a geographic information system‐based hydrologic model, was modified to represent the dominant hydrologic processes for the Palouse region of northern Idaho. Modifications included K_s as a double exponential function of depth in a single soil layer, a snow accumulation and melt algorithm, and a simple relationship between storage and perched water depth (PWD) using the drainable porosity. The model was applied to a 2 ha catchment without calibration to measured data. Distributed responses were compared with observed PWD over a 3‐year period on a 10 m × 15 m grid. Integrated responses were compared with observed surface runoff at the catchment outlet. The modified SMR model simulated the PWD fluctuations remarkably well, especially considering the shallow soils in this catchment: a 0·20 m error in PWD is equivalent to only a 1·6% error in predicted soil moisture content. Simulations also captured PWD fluctuations during a year with high spatial variability of snow accumulation and snowmelt rates at upslope, mid‐slope, and toe slope positions with errors as low as 0·09 m, 0·12 m, and 0·12 m respectively. Errors in distributed and integrated model simulations were attributed mostly to misrepresentation of rain events and snowmelt timing problems. In one location in the catchment, simulated PWD was consistently greater than observed PWD, indicating a localized recharge zone, which was not identified by the soil morphological survey. Copyright © 2006 John Wiley & Sons, Ltd.     

Uncertainty forecast from 3-D super-ensemble multi-model combination: validation and calibration

Measurements collected during the Recognized Environmental Picture 2010 experiment (REP10) in the Ligurian Sea are used to evaluate 3-D super-ensemble (3DSE) 72-hour temperature predictions and their associated uncertainty. The 3DSE reduces the total Root-Mean-Square Difference by 12 and 32% respectively with reference to the ensemble mean and the most accurate of the models when comparing to regularly distributed surface temperature data. When validating against irregularly distributed in situ observations, the 3DSE, ensemble mean and most accurate model lead to similar scores. The 3DSE temperature uncertainty estimate is obtained from the product of a posteriori model weight error covariances by an operator containing model forecast values. This uncertainty prediction is evaluated using a criterion based on the 2.5th and 97.5th percentiles of the error distribution. The 3DSE error is found to be on average underestimated during the forecast period, reflecting (i) the influence of ocean dynamics and (ii) inaccuracies in the a priori weight error correlations. A calibration of the theoretical 3DSE uncertainty is proposed for the REP10 scenario, based on a time-evolving amplification coefficient applied to the a posteriori weight error covariance matrix. This calibration allows the end-user to be confident that, on average, the true ocean state lies in the −2/+2 3DSE uncertainty range in 95% of the cases.     

On the use of multi‐algorithm,genetically adaptive multi‐objective method for multi‐site calibration of the SWAT model

With the availability of spatially distributed data, distributed hydrologic models are increasingly used for simulation of spatially varied hydrologic processes to understand and manage natural and human activities that affect watershed systems. Multi‐objective optimization methods have been applied to calibrate distributed hydrologic models using observed data from multiple sites. As the time consumed by running these complex models is increasing substantially, selecting efficient and effective multi‐objective optimization algorithms is becoming a nontrivial issue. In this study, we evaluated a multi‐algorithm, genetically adaptive multi‐objective method (AMALGAM) for multi‐site calibration of a distributed hydrologic model—Soil and Water Assessment Tool (SWAT), and compared its performance with two widely used evolutionary multi‐objective optimization (EMO) algorithms (i.e. Strength Pareto Evolutionary Algorithm 2 (SPEA2) and Non‐dominated Sorted Genetic Algorithm II (NSGA‐II)). In order to provide insights into each method's overall performance, these three methods were tested in four watersheds with various characteristics. The test results indicate that the AMALGAM can consistently provide competitive or superior results compared with the other two methods. The multi‐method search framework of AMALGAM, which can flexibly and adaptively utilize multiple optimization algorithms, makes it a promising tool for multi‐site calibration of the distributed SWAT. For practical use of AMALGAM, it is suggested to implement this method in multiple trials with relatively small number of model runs rather than run it once with long iterations. In addition, incorporating different multi‐objective optimization algorithms and multi‐mode search operators into AMALGAM deserves further research. Copyright © 2009 John Wiley & Sons, Ltd.     

A hybrid optimization approach for efficient calibration of computationally intensive hydrological models

ABSTRACT

The calibration of hydrological models is formulated as a blackbox optimization problem where the only information available is the objective function value. Distributed hydrological models are generally computationally intensive, and their calibration may require several hours or days which can be an issue for many operational contexts. Different optimization algorithms have been developed over the years and exhibit different strengths when applied to the calibration of computationally intensive hydrological models. This paper shows how the dynamically dimensioned search (DDS) and the mesh adaptive direct search (MADS) algorithms can be combined to significantly reduce the computational time of calibrating distributed hydrological models while ensuring robustness and stability regarding the final objective function values. Five transitional features are described to adequately merge both algorithms. The hybrid approach is applied to the distributed and computationally intensive HYDROTEL model on three different river basins located in Québec (Canada).     

Improving flow forecasting by error correction modelling in altered catchment conditions

Despite human is an increasingly significant component of the hydrologic cycle in many river basins, most hydrologic models are still developed to accurately reproduce the natural processes and ignore the effect of human activities on the watershed response. This results in non‐stationary model forecast errors and poor predicting performance every time these models are used in non‐pristine watersheds. In the last decade, the representation of human activities in hydrological models has been extensively studied. However, mathematical models integrating the human and the natural dimension are not very common in hydrological applications and nearly unknown in the day‐to‐day practice. In this paper, we propose a new simple data‐driven flow forecast correction method that can be used to simultaneously tackle forecast errors from structural, parameter and input uncertainty, and errors that arise from neglecting human‐induced alterations in conceptual rainfall–runoff models. The correction system is composed of two layers: (i) a classification system that, based on the current flow condition, detects whether the source of error is natural or human induced and (ii) a set of error correction models that are alternatively activated, each tailored to the specific source of errors. As a case study, we consider the highly anthropized Aniene river basin in Italy, where a flow forecasting system is being established to support the operation of a hydropower dam. Results show that, even by using very basic methods, namely if‐then classification rules and linear correction models, the proposed methodology considerably improves the forecasting capability of the original hydrological model. Copyright © 2013 John Wiley & Sons, Ltd.     

Application of the patient rule induction method to detect hydrologic model behavioural parameters and quantify uncertainty

Finding an operational parameter vector is always challenging in the application of hydrologic models, with over‐parameterization and limited information from observations leading to uncertainty about the best parameter vectors. Thus, it is beneficial to find every possible behavioural parameter vector. This paper presents a new methodology, called the patient rule induction method for parameter estimation (PRIM‐PE), to define where the behavioural parameter vectors are located in the parameter space. The PRIM‐PE was used to discover all regions of the parameter space containing an acceptable model behaviour. This algorithm consists of an initial sampling procedure to generate a parameter sample that sufficiently represents the response surface with a uniform distribution within the “good‐enough” region (i.e., performance better than a predefined threshold) and a rule induction component (PRIM), which is then used to define regions in the parameter space in which the acceptable parameter vectors are located. To investigate its ability in different situations, the methodology is evaluated using four test problems. The PRIM‐PE sampling procedure was also compared against a Markov chain Monte Carlo sampler known as the differential evolution adaptive Metropolis (DREAM_ZS) algorithm. Finally, a spatially distributed hydrological model calibration problem with two settings (a three‐parameter calibration problem and a 23‐parameter calibration problem) was solved using the PRIM‐PE algorithm. The results show that the PRIM‐PE method captured the good‐enough region in the parameter space successfully using 8 and 107 boxes for the three‐parameter and 23‐parameter problems, respectively. This good‐enough region can be used in a global sensitivity analysis to provide a broad range of parameter vectors that produce acceptable model performance. Moreover, for a specific objective function and model structure, the size of the boxes can be used as a measure of equifinality.     

Performance of a conceptual and physically based model in simulating the response of a semi‐urbanized watershed in San Antonio,Texas

The need for accurate hydrologic analysis and rainfall–runoff modelling tools has been rapidly increasing because of the growing complexity of operational hydrologic and hydraulic problems associated with population growth, rapid urbanization and expansion of agricultural activities. Given the recent advances in remote sensing of physiographic features and the availability of near real‐time precipitation products, rainfall–runoff models are expected to predict runoff more accurately. In this study, we compare the performance and implementation requirements of two rainfall–runoff models for a semi‐urbanized watershed. One is a semi‐distributed conceptual model, the Hydrologic Engineering Center‐Hydrologic Modelling System (HEC‐HMS). The other is a physically based, distributed‐parameter hydrologic model, the Gridded Surface Subsurface Hydrologic Analysis (GSSHA). Four flood events that took place on the Leon Creek watershed, a sub‐watershed of the San Antonio River basin in Texas, were used in this study. The two models were driven by the Multisensor Precipitation Estimator radar products. One event (in 2007) was used for HEC‐HMS and GSSHA calibrations. Two events (in 2004 and 2007) were used for further calibration of HEC‐HMS. Three events (in 2002, 2004 and 2010) were used for model validation. In general, the physically based, distributed‐parameter model performed better than the conceptual model and required less calibration. The two models were prepared with the same minimum required input data, and the effort required to build the two models did not differ substantially. Copyright © 2012 John Wiley & Sons, Ltd.     

Uncertainty assessment and optimization of hydrological model with the Shuffled Complex Evolution Metropolis algorithm: an application to artificial neural network rainfall-runoff model

Assessment of parameter and predictive uncertainty of hydrologic models is an essential part in the field of hydrology. However, during the past decades, research related to hydrologic model uncertainty is mostly done with conceptual models. As is accepted that uncertainty in model predictions arises from measurement errors associated with the system input and output, from model structural errors and from problems with parameter estimation. Unfortunately, non-conceptual models, such as black-box models, also suffer from these problems. In this paper, we take the artificial neural network (ANN) rainfall-runoff model as an example, and the Shuffled Complex Evolution Metropolis algorithm (SCEM-UA) is employed to analysis the parameter and predictive uncertainty of this model. Furthermore, based on the results of uncertainty assessment, we finally arrive at a simpler incomplete-connection artificial neural network (ICANN) model as well as with better performance compared to original ANN rainfall-runoff model. These results not only indicate that SCEM-UA can be a useful tool for uncertainty analysis of ANN model, but also prove that uncertainty does exist in ANN rainfall-runoff model. Additionally, in some way, it presents that the ICANN model is with smaller uncertainty than the original ANN model.     

Impacts of uncertain river flow data on rainfall‐runoff model calibration and discharge predictions

In order to quantify total error affecting hydrological models and predictions, we must explicitly recognize errors in input data, model structure, model parameters and validation data. This paper tackles the last of these: errors in discharge measurements used to calibrate a rainfall‐runoff model, caused by stage–discharge rating‐curve uncertainty. This uncertainty may be due to several combined sources, including errors in stage and velocity measurements during individual gaugings, assumptions regarding a particular form of stage–discharge relationship, extrapolation of the stage–discharge relationship beyond the maximum gauging, and cross‐section change due to vegetation growth and/or bed movement. A methodology is presented to systematically assess and quantify the uncertainty in discharge measurements due to all of these sources. For a given stage measurement, a complete PDF of true discharge is estimated. Consequently, new model calibration techniques can be introduced to explicitly account for the discharge error distribution. The method is demonstrated for a gravel‐bed river in New Zealand, where all the above uncertainty sources can be identified, including significant uncertainty in cross‐section form due to scour and re‐deposition of sediment. Results show that rigorous consideration of uncertainty in flow data results in significant improvement of the model's ability to predict the observed flow. Copyright © 2010 John Wiley & Sons, Ltd.