Bayesian updating of flood inundation likelihoods conditioned on flood extent data

Previously we have detailed an application of the generalized likelihood uncertainty estimation (GLUE) procedure to estimate spatially distributed uncertainty in models conditioned against binary pattern data contained in flood inundation maps. This method was applied to two sites where a single consistent synoptic image of inundation extent was available to test the simulation performance of the method. In this paper, we extend this to examine the predictive performance of the method for a reach of the River Severn, west‐central England. Uniquely for this reach, consistent inundation images of two major floods have been acquired from spaceborne synthetic aperture radars, as well as a high‐resolution digital elevation model derived using laser altimetry. These data thus allow rigorous split sample testing of the previous GLUE application. To achieve this, Monte Carlo analyses of parameter uncertainty within the GLUE framework are conducted for a typical hydraulic model applied to each flood event. The best 10% of parameter sets identified in each analysis are then used to map uncertainty in flood extent predictions using the method previously proposed for both an independent validation data set and a design flood. Finally, methods for combining the likelihood information derived from each Monte Carlo ensemble are examined to determine whether this has the potential to reduce uncertainty in spatially distributed measures of flood risk for a design flood. The results show that for this reach and these events, the method previously established is able to produce sharply defined flood risk maps that compare well with observed inundation extent. More generally, we show that even single, poor‐quality inundation extent images are useful in constraining hydraulic model calibrations and that values of effective friction parameters are broadly stationary between the two events simulated, most probably reflecting their similar hydraulics. Copyright © 2004 John Wiley & Sons, Ltd.     

Advances in pan‐European flood hazard mapping

Flood hazard maps at trans‐national scale have potential for a large number of applications ranging from climate change studies, reinsurance products, aid to emergency operations for major flood crisis, among others. However, at continental scales, only few products are available, due to the difficulty of retrieving large consistent data sets. Moreover, these are produced at relatively coarse grid resolution, which limits their applications to qualitative assessments. At finer resolution, maps are often limited to country boundaries, due to limited data sharing at trans‐national level. The creation of a European flood hazard map would currently imply a collection of scattered regional maps, often lacking mutual consistency due to the variety of adopted approaches and quality of the underlying input data. In this work, we derive a pan‐European flood hazard map at 100 m resolution. The proposed approach is based on expanding a literature cascade model through a physically based approach. A combination of distributed hydrological and hydraulic models was set up for the European domain. Then, an observed meteorological data set is used to derive a long‐term streamflow simulation and subsequently coherent design flood hydrographs for a return period of 100 years along the pan‐European river network. Flood hydrographs are used to simulate areas at risk of flooding and output maps are merged into a pan‐European flood hazard map. The quality of this map is evaluated for selected areas in Germany and United Kingdom against national/regional hazard maps. Despite inherent limitations and model resolution issues, simulated maps are in good agreement with reference maps (hit rate between 59% and 78%, critical success index between 43% and 65%), suggesting strong potential for a number of applications at the European scale. Copyright © 2013 John Wiley & Sons, Ltd.     

Uncertainty in floodplain delineation: expression of flood hazard and risk in a Gulf Coast watershed

This paper investigates the development of flood hazard and flood risk delineations that account for uncertainty as improvements to standard floodplain maps for coastal watersheds. Current regulatory floodplain maps for the Gulf Coastal United States present 1% flood hazards as polygon features developed using deterministic, steady‐state models that do not consider data uncertainty or natural variability of input parameters. Using the techniques presented here, a standard binary deterministic floodplain delineation is replaced with a flood inundation map showing the underlying flood hazard structure. Additionally, the hazard uncertainty is further transformed to show flood risk as a spatially distributed probable flood depth using concepts familiar to practicing engineers and software tools accepted and understood by regulators. A case study of the proposed hazard and risk assessment methodology is presented for a Gulf Coast watershed, which suggests that storm duration and stage boundary conditions are important variable parameters, whereas rainfall distribution, storm movement, and roughness coefficients contribute less variability. The floodplain with uncertainty for this coastal watershed showed the highest variability in the tidally influenced reaches and showed little variability in the inland riverine reaches. Additionally, comparison of flood hazard maps to flood risk maps shows that they are not directly correlated, as areas of high hazard do not always represent high risk. Copyright © 2012 John Wiley & Sons, Ltd.     

Assessing the uncertainty in distributed model predictions using observed binary pattern information within GLUE

In this paper we extend the generalized likelihood uncertainty estimation (GLUE) technique to estimate spatially distributed uncertainty in models conditioned against binary pattern data contained in flood inundation maps. Untransformed binary pattern data already have been used within GLUE to estimate domain‐averaged (zero‐dimensional) likelihoods, yet the pattern information embedded within such sources has not been used to estimate distributed uncertainty. Where pattern information has been used to map distributed uncertainty it has been transformed into a continuous function prior to use, which may introduce additional errors. To solve this problem we use here ‘raw’ binary pattern data to define a zero‐dimensional global performance measure for each simulation in a Monte Carlo ensemble. Thereafter, for each pixel of the distributed model we evaluate the probability that this pixel was inundated. This probability is then weighted by the measure of global model performance, thus taking into account how well a given parameter set performs overall. The result is a distributed uncertainty measure mapped over real space. The advantage of the approach is that it both captures distributed uncertainty and contains information on global likelihood that can be used to condition predictions of further events for which observed data are not available. The technique is applied to the problem of flood inundation prediction at two test sites representing different hydrodynamic conditions. In both cases, the method reveals the spatial structure in simulation uncertainty and simultaneously enables mapping of flood probability predicted by the model. Spatially distributed uncertainty analysis is shown to contain information over and above that available from global performance measures. Overall, the paper highlights the different types of information that may be obtained from mappings of model uncertainty over real and n‐dimensional parameter spaces. Copyright © 2002 John Wiley & Sons, Ltd.     

Estimation of uncertainty propagation in flood inundation mapping using a 1‐D hydraulic model

A need for more accurate flood inundation maps has recently arisen because of the increasing frequency and extremity of flood events. The accuracy of flood inundation maps is determined by the uncertainty propagated from all of the variables involved in the overall process of flood inundation modelling. Despite our advanced understanding of flood progression, it is impossible to eliminate the uncertainty because of the constraints involving cost, time, knowledge, and technology. Nevertheless, uncertainty analysis in flood inundation mapping can provide useful information for flood risk management. The twin objectives of this study were firstly to estimate the propagated uncertainty rates of key variables in flood inundation mapping by using the first‐order approximation method and secondly to evaluate the relative sensitivities of the model variables by using the Hornberger–Spear–Young (HSY) method. Monte Carlo simulations using the Hydrologic Engineering Center's River Analysis System and triangle‐based interpolation were performed to investigate the uncertainty arising from discharge, topography, and Manning's n in the East Fork of the White River near Seymour, Indiana, and in Strouds Creek in Orange County, North Carolina. We found that the uncertainty of a single variable is propagated differently to the flood inundation area depending on the effects of other variables in the overall process. The uncertainty was linearly/nonlinearly propagated corresponding to valley shapes of the reaches. In addition, the HSY sensitivity analysis revealed the topography of Seymour reach and the discharge of Strouds Creek to be major contributors to the change of flood inundation area. Copyright © 2014 John Wiley & Sons, Ltd.     

Flood hazard mapping using synthesis hydraulic and geomorphic properties at watershed scale

The occurrence of natural phenomena such as floods has caused serious consequences for human societies. The simulation of flood hazard maps and its depth in a river is one of the most complex processes in hydrology. In fact both geomorphological and hydraulic procedures for deriving the flood hazard maps and depth are imperfect at watershed scale. In this study, a combination of both procedures, using a probabilistic approach is used. Flood inundation maps for 2-, 10-, 25-,50- and 100-return period floods using flood routine within HEC-RAS in combination of Arc-GIS and topographic wetness index (TWI) map were produced. TWI threshold was identified using a maximum likelihood method in order to produce flood prone areas and calibrated over the reach of Zirab City. The correlation between TWI threshold and the flood depth was carried out and simple linear regression developed for various return periods. The resulting regression model is used in order to create flood hazard maps with various return periods at watershed scale.     

When does spatial resolution become spurious in probabilistic flood inundation predictions?

Advances in remote sensing have enabled hydraulic models to run at fine scale resolutions, producing precise flood inundation predictions. However, running models at finer resolutions increase their computational expense, reducing the feasibility of running the multiple model realizations required to undertake uncertainty analysis. Furthermore, it is possible that precision gained by running fine scale models is smoothed out when treating models probabilistically. The aim of this paper is to determine the level of spatial complexity that is required when making probabilistic flood inundation predictions. The Imera basin, Sicily is used as a case study to assess how changing the spatial resolution of the hydraulic model LISFLOOD‐FP impacts on the skill of conditional probabilistic flood inundation maps given model parameter and boundary condition uncertainties. We find that model performance deteriorates at resolutions coarser than 50 m. This is predominantly caused by changes in flow pathways at coarser resolutions which lead to non‐stationarity in the optimum model parameters at different spatial resolutions. However, although it is still possible to produce probabilistic flood maps that contain a coherent outline of the flood extent at coarser resolutions, the reliability of these maps deteriorates at resolutions coarser than 100 m. Additionally, although the rejection of non‐behavioural models reduces the uncertainty in probabilistic flood maps the reliability of these maps is also reduced. Models with resolutions finer than 50 m offer little gain in performance yet are more than an order of magnitude computationally expensive which can become infeasible when undertaking probabilistic analysis. Furthermore, we show that using deterministic, high‐resolution flood maps can lead to a spurious precision that would be misleading and not representative of the overall uncertainties that are inherent in making inundation predictions. Copyright © 2015 The Authors Hydrological Processes Published by John Wiley & Sons Ltd.     

Flood hazard and risk analysis in the southwest region of Bangladesh

Flood hazard and risk assessment was conducted to identify the priority areas in the southwest region of Bangladesh for flood mitigation. Simulation of flood flow through the Gorai and Arial Khan river system and its floodplains was done by using a hydrodynamic model. After model calibration and verification, the model was used to simulate the flood flow of 100‐year return period for a duration of four months. The maximum flooding depths at different locations in the rivers and floodplains were determined. The process in determining long flooding durations at every grid point in the hydrodynamic model is laborious and time‐consuming. Therefore the flood durations were determined by using satellite images of the observed flood in 1988, which has a return period close to 100 years. Flood hazard assessment was done considering flooding depth and duration. By dividing the study area into smaller land units for hazard assessment, the hazard index and the hazard factor for each land unit for depth and duration of flooding were determined. From the hazard factors of the land units, a flood hazard map, which indicates the locations of different categories of hazard zones, was developed. It was found that 54% of the study area was in the medium hazard zone, 26% in the higher hazard zone and 20% in the lower hazard zone. Due to lack of sufficient flood damage data, flood damage vulnerability is simply considered proportional to population density. The flood risk factor of each land unit was determined as the product of the flood hazard factor and the vulnerability factor. Knowing the flood risk factors for the land units, a flood risk map was developed based on the risk factors. These maps are very useful for the inhabitants and floodplain management authorities to minimize flood damage and loss of human lives. Copyright © 2005 John Wiley & Sons, Ltd.     

Quantifying the relative impact of hydrological and hydraulic modelling parameterizations on uncertainty of inundation maps

ABSTRACT

Flood risk management strongly relies on inundation models for river basin zoning in flood-prone and risk-free areas. Floodplain zoning is significantly affected by the diverse and concurrent uncertainties that characterize the modelling chain used for producing inundation maps. In order to quantify the relative impact of the uncertainties linked to a lumped hydrological (rainfall–runoff) model and a FLO-2D hydraulic model, a Monte Carlo procedure is proposed in this work. The hydrological uncertainty is associated with the design rainfall estimation method, while the hydraulic model uncertainty is associated with roughness parameterization. This uncertainty analysis is tested on the case study of the Marta coastal catchment in Italy, by comparing the different frequency, extent and depth of inundation simulations associated with varying rainfall forcing and/or hydraulic model roughness realizations. The results suggest a significant predominance of the hydrological uncertainty with respect to the hydraulic one on the overall uncertainty associated with the simulated inundation maps.     

Sensitivity of a hydraulic model to channel erosion uncertainty during extreme flooding

Recent research into flood modelling has primarily concentrated on the simulation of inundation flow without considering the influences of channel morphology. River channels are often represented by a simplified geometry that is implicitly assumed to remain unchanged during flood simulations. However, field evidence demonstrates that significant morphological changes can occur during floods to mobilize the boundary sediments. Despite this, the effect of channel morphology on model results has been largely unexplored. To address this issue, the impact of channel cross‐section geometry and channel long‐profile variability on flood dynamics is examined using an ensemble of a 1D–2D hydraulic model (LISFLOOD‐FP) of the ~1 : 2000 year recurrence interval floods in Cockermouth, UK, within an uncertainty framework. A series of simulated scenarios of channel erosional changes were constructed on the basis of a simple velocity‐based model of critical entrainment. A Monte‐Carlo simulation framework was used to quantify the effects of this channel morphology together with variations in the channel and floodplain roughness coefficients, grain size characteristics and critical shear stress on measures of flood inundation. The results showed that the bed elevation modifications generated by the simplistic equations reflected an approximation of the observed patterns of spatial erosion that enveloped observed erosion depths. The effect of uncertainty on channel long‐profile variability only affected the local flood dynamics and did not significantly affect the friction sensitivity and flood inundation mapping. The results imply that hydraulic models generally do not need to account for within event morphodynamic changes of the type and magnitude of event modelled, as these have a negligible impact that is smaller than other uncertainties, e.g. boundary conditions. Instead, morphodynamic change needs to happen over a series of events to become large enough to change the hydrodynamics of floods in supply limited gravel‐bed rivers such as the one used in this research. Copyright © 2014 John Wiley & Sons, Ltd.     

HAND contour: a new proxy predictor of inundation extent

Tools for accurately predicting environmental risks, such as the location and spatial extent of potential inundation, are not widely available. A dependence on calibration and a lack of available flood data have prevented the widespread application of existing hydrodynamic methods for predicting the extent of inundation. We use the height above the nearest drainage (HAND) terrain model to develop a simple static approach for mapping the potential extent of inundation that does not depend on flood observations and extends beyond methods for mapping low‐lying areas. While relying on the contour concept, the method utilizes drainage‐normalized topography and flowpaths to delineate the relative vertical distances (drop) to the nearest river. The HAND‐delineated relative drop is an effective distributed predictor of flood potential, which is directly related to the river stage height. We validated the new HAND contour approach using a flood event in Southern Brazil for which high‐resolution maps were available. The results indicated that the flood hazard‐mapping method accurately predicted the inundation extent of the channel carrying the flood wave and the channels influenced by flooding. For channels positioned outside of the flood‐wave area, the method overestimated the actual flood extent. As an original static assessment of floodwaters across the landscape, the HAND contour method could be used to map flood hazards in areas with poor information and could promote the development of new methods for predicting hydrological hazards. Copyright © 2015 John Wiley & Sons, Ltd.     

Investigating the influence of minor hydraulic structures on modeling flood events in lowland areas

Flood inundation models have been recognized to be a valuable tool to reproduce flow dynamics in a given area and support decision‐making processes on flood management measures. In many cases, in the simulation of flood events, only the main river channel and the associated structures are represented within the model. However, during flood events involving lowland areas, the minor drainage network – and the associated hydraulic structures – may have an important role in conveying flow and determining which areas will be flooded. The objective of this study is to investigate whether – and to what extent – small hydraulic structures in drainage networks have an influence in flooding on lowland areas. The case study for this research is the 1990 flood event which occurred in the lowland plain of the Reno River, in Northern Italy. The study area is mainly used for agricultural purposes and has a drainage system with several small bridges and culverts. The influence of the minor hydraulic structures on flood dynamics was analyzed through a combined use of one‐dimensional (1D) and two‐dimensional (2D) hydraulic models. First, a number of detailed and simplified approaches to represent hydraulic structures in the computational grids were analyzed by means of the HECRAS 1D model. Second, these approaches were implemented and tested in several 2D simulations of the flood event. The simulated inundation extents and flood levels were then compared with the observed data and with each other. The analysis of results showed that simplified schematizations were sufficient to obtain good model predictions of peak inundation extent and flood levels, at least for the present case study. Moreover, the influence of the structures on the peak flood inundation extent and flood levels was found to be limited, whereas it showed to be more significant during the drainage phase of the flood. Copyright © 2013 John Wiley & Sons, Ltd.     

Copula‐based flood frequency (COFF) analysis at the confluences of river systems

Many civil infrastructures are located near the confluence of two streams, where they may be subject to inundation by high flows from either stream or both. These infrastructures, such as highway bridges, are designed to meet specified performance objectives for floods of a specified return period (e.g. the 100 year flood). Because the flooding of structures on one stream can be affected by high flows on the other stream, it is important to know the relationship between the coincident exceedence probabilities on the confluent stream pair in many hydrological engineering practices. Currently, the National Flood Frequency Program (NFF), which was developed by the US Geological Survey (USGS) and based on regional analysis, is probably the most popular model for ungauged site flood estimation and could be employed to estimate flood probabilities at the confluence points. The need for improved infrastructure design at such sites has motivated a renewed interest in the development of more rigorous joint probability distributions of the coincident flows. To accomplish this, a practical procedure is needed to determine the crucial bivariate distributions of design flows at stream confluences. In the past, the copula method provided a way to construct multivariate distribution functions. This paper aims to develop the Copula‐based Flood Frequency (COFF) method at the confluence points with any type of marginal distributions via the use of Archimedean copulas and dependent parameters. The practical implementation was assessed and tested against the standard NFF approach by a case study in Iowa's Des Moines River. Monte Carlo simulations proved the success of the generalized copula‐based joint distribution algorithm. Copyright © 2009 John Wiley & Sons, Ltd.     

东南沿海水库下游地区基于动态模拟的洪涝风险评估

我国东南沿海地区大多为一些中小流域,这些流域上游多建有水库工程,下游则为人口稠密的平原区,流域调蓄能力小,汇流时间短.同时,随着近年来城镇化快速发展,洪涝风险不断加大.因此,迫切需要开展水库下游不同暴雨重现期下的洪涝风险评估研究,以便为防洪决策提供技术支撑.为此,本文利用遥感、GIS、水文水动力学模型等相关技术方法,建立洪涝动态模拟模型来评估洪涝危险性;采用层次分析法和因子叠加法,从洪涝危险性和洪涝易损性两方面开展洪涝风险综合评估分析.研究表明,通过多学科与多技术手段相结合方法,来模拟预测不同暴雨重现期洪水动态淹没过程,再结合相关社会经济属性,可以有效地评估研究区洪涝灾害的风险,从而为水库调度及流域防洪减灾提供有力支撑.     

Numerical simulation of overbank processes in topographically complex floodplain environments

This article presents results from an investigation of the hydraulic characteristics of overbank flows on topographically‐complex natural river floodplains. A two‐dimensional hydraulic model that solves the depth‐averaged shallow water form of the Navier–Stokes equations is used to simulate an overbank flow event within a multiple channel reach of the River Culm, Devon, UK. Parameterization of channel and floodplain roughness by the model is evaluated using monitored records of main channel water level and point measurements of floodplain flow depth and unit discharge. Modelled inundation extents and sequences are assessed using maps of actual inundation patterns obtained using a Global Positioning System, observational evidence and ground photographs. Simulation results suggest a two‐phase model of flooding at the site, which seems likely to be representative of natural floodplains in general. Comparison of these results with previous research demonstrates the complexity of overbank flows on natural river floodplains and highlights the limitations of laboratory flumes as an analogue for these environments. Despite this complexity, frequency distributions of simulated depth, velocity and unit discharge data closely follow a simple gamma distribution model, and are described by a shape parameter (α) that exhibits clear systematic trends with changing discharge and floodplain roughness. Such statistical approaches have the potential to provide the basis for computationally efficient flood routing and overbank sedimentation models. Copyright © 2002 John Wiley & Sons, Ltd.     

A steepness index unit volume flood hydrograph approach for sub‐daily flow disaggregation

The estimation of sub‐daily flows from daily flood flows is important for many hydrological and hydraulic applications. Flows during flood events often vary significantly within sub‐daily time‐scales, and failure to capture the sub‐daily flood characteristic can result in an underestimation of the instantaneous flood peaks, with possible risk of design failure. It is more common to find a longer record of daily flow series (observed or modelled using daily rainfall series) than sub‐daily flow data. This paper describes a novel approach, known as the steepness index unit volume flood hydrograph approach, for disaggregating daily flood flows into sub‐daily flows that takes advantage of the strong relationship between the standardized instantaneous flood peak and the standardized daily flood hydrograph rising‐limb steepness index. The strength of this relationship, which is considerably stronger than the relationship between the standardized flood peak and the event flood volume, is shown using data from six rivers flowing into the Gippsland Lakes in southeast Australia. The results indicate that the steepness index unit volume flood hydrograph approach can be used to disaggregate modelled daily flood flows satisfactorily, but its reliability is dependent on a model's ability to simulate the standardized daily flood hydrograph rising‐limb steepness index and the event flood volume. Copyright © 2007 John Wiley & Sons, Ltd.     

A fast simplified model for predicting river flood inundation probabilities in poorly gauged areas

The capability of a simple kinematic‐storage model (KSM) is analysed to be used as a tool for a Decision Support System for the evaluation of probability inundation maps in near real time in poorly gauged areas. KSM simulates the floodplain as a storage and assumes no exchange of momentum with the channel. For the in‐bank flow, the storage is modified through a coefficient for taking the variations of channel cross sections into account. The generalized likelihood uncertainty estimation approach is used for addressing the probability flood maps along with their associated uncertainties. The model is tested on two river reaches along the Tiber River in Central Italy where observed inundation maps are available for two recent flood events. Despite the inherent uncertainties present in the input data and in the model structure, the results show that the model reproduces reasonably well, in terms of both precision and accuracy, the observed inundated areas. Tests were performed at different digital elevation model resolutions, showing a small effect of the geometry on the maximum performance obtained. The very low computational times, the parsimony of the model and its low sensitivity to the quality of the geometry representation of the channel and the floodplain makes KSM very appealing for flood forecasting and early warning system applications in poorly gauged and inaccessible areas. Copyright © 2014 John Wiley & Sons, Ltd.     

A modelling framework for evaluation of the hydrological impacts of nature‐based approaches to flood risk management,with application to in‐channel interventions across a 29‐km2 scale catchment in the United Kingdom

Nature‐based approaches to flood risk management are increasing in popularity. Evidence for the effectiveness at the catchment scale of such spatially distributed upstream measures is inconclusive. However, it also remains an open question whether, under certain conditions, the individual impacts of a collection of flood mitigation interventions could combine to produce a detrimental effect on runoff response. A modelling framework is presented for evaluation of the impacts of hillslope and in‐channel natural flood management interventions. It couples an existing semidistributed hydrological model with a new, spatially explicit, hydraulic channel network routing model. The model is applied to assess a potential flood mitigation scheme in an agricultural catchment in North Yorkshire, United Kingdom, comprising various configurations of a single variety of in‐channel feature. The hydrological model is used to generate subsurface and surface fluxes for a flood event in 2012. The network routing model is then applied to evaluate the response to the addition of up to 59 features. Additional channel and floodplain storage of approximately 70,000 m³ is seen with a reduction of around 11% in peak discharge. Although this might be sufficient to reduce flooding in moderate events, it is inadequate to prevent flooding in the double‐peaked storm of the magnitude that caused damage within the catchment in 2012. Some strategies using features specific to this catchment are suggested in order to improve the attenuation that could be achieved by applying a nature‐based approach.     

How much physical complexity is needed to model flood inundation?

Two‐dimensional flood inundation models are widely used tools for flood hazard mapping and an essential component of statutory flood risk management guidelines in many countries. Yet, we still do not know how much physical complexity a flood inundation model needs for a given problem. Here, three two‐dimensional explicit hydraulic models, which can be broadly defined as simulating diffusive, inertial or shallow water waves, have been benchmarked using test cases from a recent Environment Agency for England and Wales study, where results from industry models are also available. To ensure consistency, the three models were written in the same code and share subroutines for all but the momentum (flow) and time‐stepping calculations. The diffusive type model required much longer simulation times than the other models, whilst the inertia model was the quickest. For flows that vary gradually in time, differences in simulated velocities and depths due to physical complexity were within 10% of the simulations from a range of industry models. Therefore, for flows that vary gradually in time, it appears unnecessary to solve the full two‐dimensional shallow water equations. As expected, however, the simpler models were unable to simulate supercritical flows accurately. Finally, implications of the results for future model benchmarking studies are discussed in light of a number of subtle factors that were found to cause significant differences in simulations relative to the choice of model. Copyright © 2011 John Wiley & Sons, Ltd.