211 research outputs found
Sensitivity analysis of data-driven groundwater forecasts to hydroclimatic controls in irrigated croplands
No full textIn the last decades, advancements in computational science have greatly expanded the use of artificial neural networks (ANNs) in hydrogeology, including applications on groundwater forecast, variable selection, extended lead-times, and regime-specific analysis. However, ANN-model performance often omits the sensitivity to observational uncertainties in hydroclimate forcings. The goal of this paper is to implement a data-driven modeling framework for assessing the sensitivity of ANN-based groundwater forecasts to the uncertainties in observational inputs across space, time, and hydrological regimes. The objectives are two-folded. The first objective is to couple an ANN model with the PAWN sensitivity analysis (SA). The second objective is to evaluate the scale- and process-dependent sensitivities of groundwater forecasts to hydroclimate inputs, computing the sensitivity index in groundwater wells (1) across the whole time-series (for the global sensitivity analysis); (2) across the output sub-regions with conditions of water deficit and water surplus (for the ‘regional’ sensitivity analysis); and (3) at each time step (for the time-varying sensitivity analysis). The implementation of the ANN-PAWN occurs in 68 wells across the Northern High Plains aquifer, USA, with pre-time-step rainfall, evapotranspiration, snowmelt, streamflow, and groundwater measurements as inputs. Results show that evapotranspiration and rainfall are the major sources of uncertainty, with the latter being particularly relevant in water surplus conditions and the former in water deficit conditions. The time-varying sensitivity analysis leads to the identification of localized sensitivities to other sources of uncertainty, as snowmelt in spring or river flow during the annual peak period at the groundwater level
Integrating Qualitative Flow Observations in a Lumped Hydrologic Routing Model
No full textThis study aims at proposing novel approaches for integrating qualitative flow observations in a lumped hydrologic routing model and assessing their usefulness for improving flood estimation. Routing is based on a three-parameter Muskingum model used to propagate streamflow in five different rivers in the United States. Qualitative flow observations, synthetically generated from observed flow, are converted into fuzzy observations using flow characteristic for defining fuzzy classes. A model states updating method and a model output correction technique are implemented. An innovative application of Interacting Multiple Models, which use was previously demonstrated on tracking in ballistic missile applications, is proposed as state updating method, together with the traditional Kalman filter. The output corrector approach is based on the fuzzy error corrector, which was previously used for robots navigation. This study demonstrates the usefulness of integrating qualitative flow observations for improving flood estimation. In particular, state updating methods outperform the output correction approach in terms of average improvement of model performances, while the latter is found to be less sensitive to biased observations and to the definition of fuzzy sets used to represent qualitative observations.Water Resource
Committees Of Specialized Conceptual Hydrological Models: Comparative Study
No full textCommittee modelling approach is skillful prediction in the domain of hydrological modelling that allows explicitly to derive predictive model outputs. In this approach, the different individual models are optimally combined. Generally if a single hydrological model or the model calibrated by the single aggregated objective function it is hard to capture all facets of a complex process and to present the best possible model outputs. This model could be either capable for high flows or for low flows or not for both cases hence more flexible modelling architectures are required. Here the possibilities is building several specialized models each of which is responsible for a particular sub-process (high flows or low flows), and combining them using dynamic weights – thus forming a committee model. In this study we compare two different types of committee models: (i) the combine model based on fuzzy memberships function (Kayastha et al. 2013, Fenicia et al. 2007) and (ii) the combine model based on weights that calculated from hydrological states (Oudin et al. 2006). Before combining the models the individual hydrological models are calibrated by Adaptive Cluster Covering Algorithm (Solomatine 1999) for high and low flows with (different) suitable objective functions. The committee model based on fuzzy memberships does not generate additional water in the system (preserves water balance), however there is no guarantee for this in case of committees based on hydrological states. The relative performances of the two different committee models and their characteristics are illustrated, with an application to HBV hydrological models in Bagmati catchment in Nepal
Erratum for “Multiobjective Evolutionary Approach to Rehabilitation of Urban Drainage Systems” by Wilmer Barreto, Zoran Vojinovic, Roland Price, and Dimitri Solomatine
No full textMulti-objective valve management optimization formulations for water quality enhancement in WDNs.
No full textWater distribution networks (WDNs) need to guarantee that water is delivered with adequate quality. This paper compares the
performance of 12 multiobjective procedures to limit water quality deterioration in a WDN through the optimal operation of valves. The first
objective (ObF1) is to minimize the water age, chosen as a surrogate parameter of quality deterioration, and the second objective (ObF2) is to
minimize the number of valve closures. The 12 procedures are derived from the combination of 4 different optimization algorithms and
3 formulations of ObF1, namely, to minimize the maximum, the arithmetic mean, and the demand-weighted mean water age. The optimization
algorithms considered are random search (RS), Loop for Optimal Valve Status Configuration (LOC), and a combination of each
of these two with the Archive-based Micro Genetic Algorithm. The procedures are tested on two networks of different complexity. Results
show how LOC is able to find near-optimal solutions using a fraction of the computational time required by a brute force search. Furthermore,
among the ObF1 formulations, the use of the averages (either arithmetic or demand-weighted) gives better results in terms of impact on the
population served by a WDN
Robust optimization of valve management to improve water quality in WDNs under demand uncertainty.
No full textWater quality deterioration in water distribution networks can be associated with high water residence time in the network. To this end, some previous studies have proposed optimization procedures for valve management. However, these studies generally come up with operational configurations assuming deterministic user demand patterns that may never occur in reality. In consequence, the proposed solutions may not be effective for improving water quality or do not comply with pressure constraints if different demand patterns are observed. This study proposes a methodology to determine robust configurations of the valves to keep water residence time at acceptable levels regardless of the variability in demand patterns. The methodology is tested on four different distribution systems of varying topology and size. Results show the importance of executing robust – instead of deterministic, optimization to find valve configurations that guarantee the performance of the networks in terms of hydraulics and water quality
Prediction Of Hydrological Models’ Uncertainty By A Committee Of Machine Learning-Models
No full textThis study presents an approach to combine uncertainties of the hydrological model outputs predicted from a number of machine learning models. The machine learning based uncertainty prediction approach is very useful for estimation of hydrological models\u27 uncertainty in particular hydro-metrological situation in real-time application [1]. In this approach the hydrological model realizations from Monte Carlo simulations are used to build different machine learning uncertainty models to predict uncertainty (quantiles of pdf) of the a deterministic output from hydrological model . Uncertainty models are trained using antecedent precipitation and streamflows as inputs. The trained models are then employed to predict the model output uncertainty which is specific for the new input data. We used three machine learning models namely artificial neural networks, model tree, locally weighted regression to predict output uncertainties. These three models produce similar verification results, which can be improved by merging their outputs dynamically. We propose an approach to form a committee of the three models to combine their outputs. The approach is applied to estimate uncertainty of streamflows simulation from a conceptual hydrological model in the Brue catchment in UK and the Bagmati catchment in Nepal. The verification results show that merged output is better than an individual model output. [1] D. L. Shrestha, N. Kayastha, and D. P. Solomatine, and R. Price. Encapsulation of parameteric uncertainty statistics by various predictive machine learning models: MLUE method, Journal of Hydroinformatic, in press, 2013
Refining the Committee Approach and Uncertainty Prediction in Hydrological Modelling
No full textDue to the complexity of hydrological systems a single model may be unable to capture the full range of a catchment response and accurately predict the streamflows. The multi modelling approach opens up possibilities for handling such difficulties and allows improve the predictive capability of models. One of multi modelling approaches called "committee modelling" is one of the topics in part of this study. Special attention is given to the so-called “fuzzy committee” approach to hydrological modelling. The comparative interpretation of the resulting uncertainty statistics from different sampling schemes (MCS, GLUE, MCMC, SCEMUA, DREAM, PSO, and ACCO) for uncertainty estimations of hydrological model is presented. The uncertainty statistics are considerably depending on the sampling method used. Another aspect of uncertainty analysis relates to predicting uncertainty (rather than its analysis). Machine learning techniques were proposed to build model of probability distribution function as predictive uncertainty models, which allows adequate uncertainty estimation for hydrological models. In flood modelling hydrological models are typically used in combination with hydraulic models forming a cascade, often supported by geospatial processing. SWAT hydrological and SOBEK hydrodynamic models are integrated for uncertainty analysis of flood inundation modelling of the Nzoia catchment (Kenya), and the parametric uncertainty of the hydrological model is allowed to propagate through the model cascade using Monte Carlo simulations, leading to the generation of the probabilistic flood maps. Due to the high computational complexity of these experiments, the high performance (cluster) computing framework is designed and used. Overall, this thesis presents research efforts in: (i) committee modelling of hydrological models, (ii) hybrid committee hydrological models, (iii) influence of sampling strategies on prediction uncertainty of hydrological models, (iv) uncertainty prediction using machine learning techniques, (v) committee of predictive uncertainty models and (vi) uncertainty in flood inundation extent. This study refined a number of hydroinformatics techniques, thus enhancing uncertainty-based hydrological and integrated modelling.Water ManagementCivil Engineering and Geoscience
Predicting Storm Surges: Chaos, Computational Intelligence, Data Assimilation, Ensembles
No full textAccurate predictions of storm surge are of importance in many coastal areas. This book focuses on data-driven modelling using methods of nonlinear dynamics and chaos theory for predicting storm surges. A number of new enhancements are presented: phase space dimensionality reduction, incomplete time series, phase error correction, finding true neighbours, optimization of chaotic model, data assimilation and multi-model ensembles. These were tested on the case studies in the North Sea and Caribbean Sea. Chaotic models appear to be are accurate and reliable short and mid-term predictors of storm surges aimed at supporting decision-makers for flood prediction and ship navigation.Water ManagementCivil Engineering and Geoscience
Framework for Dynamic Modelling of Urban Floods at Different Topographical Resolutions
No full textFloods are among the most frequent and costly natural disasters in terms of human hardship and economic loss. The impacts of flooding are especially devastating in urban areas as these areas are densely populated and contain vital infrastructures. Urban flood risks and their impacts are expected to increase as urban development in flood prone areas continues and as rain intensity increases as a result of climate change while aging drainage infrastructures limit the drainage capacity in existing urban areas. The increased risk and severe consequence of flooding drives the need for the development of costeffective flood mitigation strategies as part of sound urban flood management plans. Efficient prediction of characteristics of flood propagation in urban areas is paramount in developing flood mitigation measures. Urban flood modelling attempts to quantitatively describe the characteristics and evolution of flood flows that occur when a large amount of water moves along drainage systems and urban flood plains.Water ManagementCivil Engineering and Geoscience
- …
