Flood Frequency Curves Research Articles

Spatially smooth regional estimation of the flood frequency curve (with uncertainty)

Identification of the flood frequency curve in ungauged basins is usually performed by means of regional models based on the grouping of data recorded at various gauging stations. The present work aims at implementing a regional procedure that overcomes some of the limitations of the standard approaches and adds a clearer representation of the uncertainty components of the estimation. The information in the sample records is summarized in a set of sample L-moments, that become the variables to be regionalized. To transfer the information to ungauged basins we adopt a regional model for each of the L-moments, based on a comprehensive multiple regression approach. The independent variables of the regression are selected among a large number of geomorpholoclimatic catchment descriptors. Each model is calibrated on the entire dataset of stations using non-standard least-squares techniques accounting for the sample variability of L-moments, without resorting to any grouping procedure to create sub-regions. In this way, L-moments are allowed to vary smoothly from site to site in the descriptor space, following the variation of the descriptors selected in the regression models. This approach overcomes the subjectivity affecting the techniques for the definition and verification of the homogeneous regions. In addition, the method provides accurate confidence bands for the frequency curves estimated in ungauged basins. The procedure has been applied to a vast region in North-Western Italy (about 30,000 km^2). Cross-validation techniques are used to assess the efficiency of this approach in reconstructing the flood frequency curves, demonstrating the feasibility and the robustness of the approach

Regional Statistical Models for the Estimation of Flood Peak Values at Ungauged Catchments: Peninsular Malaysia

Multivariate regional statistical models were developed to estimate flood peak values of various return periods in the Peninsular Malaysia. Annual maximum flood peak data were collected, screened, and analyzed to identify the river gauging stations' potential to produce identical flood frequency curves. The Peninsula was divided into 10 flood regions. The mean annual flood (MAF) for each region was considered as the function of catchment area (A) and mean annual rainfall (MAR). The regional equations exhibited good coefficient of determination (R 2 ). Except for the west of Johor Baru State (Region F6), which showed an R 2 value of 0.874, other regions had values greater than 0.90. The regions located along the eastcoast of the Peninsula exhibited better coefficients of determination compared with those for the westcoast, owing to the influence of the northeast monsoon along the eastcoast. Mean index errors (MIEs) with respect to the actual MAF of each flood region are provided to rationalize the design flood peak values to minimize the uncertainty in the predictions. Flood frequency values for the lower, mean, and upper limits were proposed to reduce the effect of outliers and uncertainty in prediction of flood peak values. Knowing the catchment area and MAR of the regions, the design flood peaks of various frequencies can be estimated for the rural ungauged catchments in the Peninsula. DOI: 10.1061/(ASCE)HE.1943-5584.0000464. © 2012 American Society of Civil Engineers. CE Database subject headings: Floods; Flood frequency; Rainfall; Catchments; Malaysia.

Design soil moisture estimation by comparing continuous and storm‐based rainfall‐runoff modeling

In recent years local, national, and international authorities have showed an increasing awareness of flood and inundation hazard, likely due to the large floods which occurred in the past years in many regions of the world. In this context, the estimation of the design flood values to be adopted for flood risk assessment or floodplain management represents a crucial factor. In the case of ungauged or scarcely gauged catchments where a sufficiently long discharge time series is missing, a relevant uncertainty is involved in the flood frequency analysis and a possible solution to reduce this uncertainty is the application of continuous simulation (CS) approaches. Because of the complex structure of this type of approaches and pursuing the parameters parsimony criteria, in the hydrological practice the approaches based on the design storm (DS) estimation are more widely known and applied, mainly for their simplicity. However, one major limit of the DS method is the choice of the “design soil moisture” conditions, representing a critical parameter for assessing the initial wetness of the basin. To that end, this study of investigating six subcatchments of the upper Tiber River basin (Central Italy), with drainage area ranging from 13 to 284 km2, proposes a procedure based on the application of the CS approach as a tool to define the design soil moisture to be afterwards incorporated into the more simple DS method. For each catchment, the procedure consists of (1) stochastic generation of long synthetic rainfall and temperature series starting from observed hourly data; (2) application of a lumped continuous rainfall‐runoff model to generate synthetic discharge series and, hence, to obtain the corresponding flood frequency curves; (3) estimation of the design soil moisture, for each return period, by varying in the DS approach the initial wetness conditions of the catchment so that the peak discharge estimated by the DS method matches the one given by the synthetic flood frequency curve. Moreover, in order to apply the more simple DS approach avoiding the use of the CS one, a preliminary analysis to regionalize the design soil moisture as a function of the geo‐morphological characteristics and the return period is also shown.

Stability of Confidence Levels for Flood Frequencies Using Additional Data

A comparison of flood values based on prior data and new additional data at various confidence levels is given for five representative sites in Los Angeles, California. The methodology uses computer simulations to give the confidence level values and flood frequency curves. These calculations show confidence level increases roughly in the range of 5% to 10% using some quarter of a century of additional data.

Paleohydrologic bounds and extreme flood frequency of the Upper Arkansas River, Colorado, USA

The Upper Arkansas River basin has experienced notable large floods, including the event of 2–6 June 1921 that devastated the city of Pueblo, Colorado. We investigated flood and paleoflood hydrology at strategic sites to determine the frequency and geographic extent of extreme floods within the basin for a dam safety application. Streamgage, historical, and paleoflood data were utilized to develop frequency curves at sites near Salida, Cotopaxi, Parkdale, and Pueblo. Soil/stratigraphic descriptions, radiocarbon dating, and hydraulic modeling were used to estimate paleoflood nonexceedance bounds at the four sites, which ranged from 400 to 2200 YBP for late Holocene surfaces to late Pleistocene surfaces near Cotopaxi. Peak-flow data are from lower-magnitude snowmelt runoff in May and June in the upper basin and from high-magnitude rainfall runoff from June to August in the lower basin. Flood frequency curves reflect this transition near Parkdale from snowmelt to extreme rainfall-runoff. For similar return periods, paleoflood peak discharges increase from about 480 m 3/s upstream at Loma Linda to about 4250 m 3/s downstream near Pueblo. This increase is attributed to the larger rainfall component derived from lower elevations between Loma Linda and Pueblo. Return periods for design floods at Pueblo Dam exceeded 10,000 years based on paleoflood frequency curve extrapolations.

Modelling the impact of urbanization on flood frequency relationships in the UK

This paper investigates the effect of urbanization on the three key statistics used to establish flood frequency curves when combining the index flood method with the method of L-moments for estimating distribution parameters, i.e. the median annual maximum peak flow (the index flood), and the high-order L-moment ratios L-CV and L-SKEW. An existing procedure employing catchment descriptors was used to estimate the three statistics at ungauged sites in the UK. As-rural estimates of the three statistics were obtained in 200 urban catchments and compared to the corresponding values obtained from observed data. The (log) differences of these estimates were related to catchment descriptors relevant to the urbanization process using linear regression. The results show that urbanization leads to a reduction in L-CV but an increase in L-SKEW. A jack-knife leave-one-out experiment showed that the adjustment factors developed were generally better at predicting the effect of urbanization on the flood frequency curve than the existing adjustment factor currently used in the UK.

Use of soil data in a grid-based hydrological model to estimate spatial variation in changing flood risk across the UK

A grid-based flow routing and runoff-production model, configured to employ as input either observed or Regional Climate Model (RCM) estimates of precipitation and potential evaporation (PE), has previously been used to assess how climate change may impact river flows across the UK. The slope-based Grid-to-Grid (G2G) model adequately simulated observed river flows under current climate conditions for high relief catchments, but was less successful when applied to lower-relief and/or groundwater-dominated areas. The model has now been enhanced to employ a soil dataset to configure the probability-distributed store controlling soil-moisture and runoff generation within each grid-cell. A comparison is made of the ability of both models to simulate gauged river flows across a range of British catchments using observations of rainfall and PE as input. Superior performance from the enhanced G2G formulation incorporating the soil dataset is demonstrated. Following the model assessment, the observed precipitation and PE data used as input to both hydrological models were replaced by RCM estimates on a 25 km grid for a Current (1961–1990) and a Future (2071–2100) time-slice. Flood frequency curves derived from the flow simulations for the two time-slices are used to estimate, for the first time, maps of changes in flood magnitude for all river points on a 1 km grid across the UK. A high degree of spatial variability is seen in the estimated change in river flows, reflecting both projected climate change and the influences of landscape and climate variability. These maps also highlight large differences between the climate impact projections arising from the two models. The improved structure and performance of the soil-based G2G model adds confidence to its projections of flow changes being realistic consequences of the climate change scenario applied. A resampling method is used to identify regions where these projections may be considered robust. However, with the climate change scenario used representing only one plausible evolution of the future climate, no clear message can be drawn here about projected river flow changes.

Integrated Ecohydrologic Research and Hydro‐Informatics

The purpose of this paper is to offer our view of where ecohydrologic research will be going in the next 20 years and suggest how enabling technologies from hydro-informatics will support this research. Two decades ago Klemes (1986) suggested that the hydrologist’s “efforts expended on the fitting of flood and drought frequency curves would be better spent in acquiring deeper knowledge of climatology, meteorology, geology, and ecology.” Klemes was of course calling for interdisciplinary hydrology. Recently, a number of community reports have proposed a more interdisciplinary approach to hydrology, including the development of community infrastructure such as large scale hydrologic observatories with integrated, multi-scale monitoring and advanced informatics tools to enable this research (Band et al. 2002, Gupta et al. 1999, Hornberger et al. 2000, Maidment 2008). Specific calls were included to integrate the more physically or statistically oriented approaches in hydrology with ecosystem sciences including biogeochemical cycling and population ecology. The emergence of ecohydrologic research is one example of how hydrologic science has begun to move in this direction. Ecohydrologic research seeks to understand how hydrological processes affect biological communities, and in turn how such communities affect water cycling (Newman et al. 2006, Rodriguez-Iturbe 2000). With this marriage of ecology and hydrology new avenues of research are opening up, and with these come new scientific and technical challenges. Some of the scientific challenges relate to the long-term memory in biological and geomorphic systems, complex feedbacks on water cycling, and the continuum of such interactions across space. Technical challenges include building more sophisticated simulation models to deal with such spatial dynamics, acquiring and managing the data needed to support these models, improving geo-visualization of spatial predictions and errors, and quantifying uncertainty associated with model structure and parameterization. Another interdisciplinary subfield of hydrology, hydroinformatics, emphasizes the development of information technology to help meet these challenges. Newman et al. (2006) identified a number of research challenges for ecohydrologic research in semi-arid regions, including dealing with spatial and temporal heterogeneity, scaling up to regional and global extent, improving understanding of subsurface processes, and addressing long-term processes. They argue for a greater emphasis on place-based research where long-term data sets are being compiled. Efforts aimed at addressing these problems are underway, albeit with a focus on vegetation in semi-arid environments, equilibrium models with stochastic inputs, and knowledge obtained in traditional plots or stands. We suggest that for ecohydrologic research to be globally relevant it must embrace the full spectrum of environments, including non-water limited regions and wetland-rich regions. Over the next two decades, ecohydrologic research will explore more deeply the nature of transient system evolution and elucidate characteristic timescales of processes, such as those associated with ecosystem aggradation and degradation. It will move toward developing predictive capability that builds from an understanding of processes along spatial gradients, including adaptations of nutrient cycling and plant hydraulics at wetland-upland transitions. Moreover, as cyber-infrastructure improves these activities will transcend individual study sites by utilizing combinations of data sets 16

Flood variability east of Australia’s Great Dividing Range

Summary The variability of flow in river channels influences the spatial and temporal variability of many biophysical processes including the transport of sediment and waterborne pollutants and the recruitment of aquatic animals and plants. In this study, inter- and intra-basin patterns of flood variability are examined for catchments east of Australia’s Great Dividing Range. Three measures of flood variability are explored with uncertainty quantified using bootstrap resampling. The two preferred measures of flood variability (namely a flood quantile ratio and a power law scaling coefficient) produced similar results. Catchments in the wet tropics of far north Queensland experience low flood variability. Flood variability increased southwards through Queensland, reaching a maximum in the vicinity of the Fitzroy and Burnett River basins. The small near-coast catchments of southern Queensland and northern New Wales experience low flood variability. Flood variability is also high in the southern Hunter River and Hawkesbury–Nepean basins. Using L-moment ratio diagrams with data from 424 streamflow stations, we also conclude that the Generalised Pareto distribution is preferable for modelling flood frequency curves for this region. These results provide a regional perspective that can be used to develop new hypotheses about the effects of hydrologic variability on the biophysical characteristics of these Australian rivers.

Use of L-Moments Approach for Regional Flood Frequency Analysis in Sicily, Italy

Extremely great floods are among environmental events with the most disastrous consequences for the entire world. Estimates of their return periods and design values are of great importance in hydrologic modeling, engineering practice for water resources and reservoirs design and management, planning for weather-related emergencies, etc. Regional flood frequency analysis resolves the problem of estimating the extreme flood events for catchments having short data records or ungauged catchments. This paper analyzes annual maximum peak flood discharge data recorded from more than 50 stream flow gauging sites in Sicily, Italy, in order to derive regional flood frequency curves. First these data are analyzed to point out some problems concerning the homogeneity of the single time series. On the basis of the L-moments and using cluster analysis techniques, the entire region is subdivided in five subregions whose homogeneity is tested using the L-moments based heterogeneity measure. Comparative regional flood frequency analysis studies are carried out employing the L-moments based commonly used frequency distributions. Based on the L-moment ratio diagram and other statistic criteria, generalized extreme value (GEV) distribution is identified as the robust distribution for the study area. Regional flood frequency relationships are developed to estimate floods at various return periods for gauged and ungauged catchments in different subregions of the Sicily. These relationships have been implemented using the L-moment based GEV distribution and a regional relation between mean annual peak flood and some geomorphologic and climatic parameters of catchments.

Flood regimes in the Southern Caucasus: the influence of precipitation on mean annual floods and frequency curves

Relations between mean annual flood estimates and basin area and precipitation, as well as dimensionless flood frequency curves, have been derived from a number of groups of gauging stations in the southern Caucasus. Total precipitation and its seasonal distribution are extremely variable at this location. The importance of antecedent soil moisture deficit (closely linked to the seasonal distribution of precipitation) in determining the shape of flood frequency curves is discussed through previous empirical studies and recent sensitivity analyses. The varying shape of regional curves from the southern Caucasus is related to the variations in soil moisture deficit.

Effects of multiscale rainfall variability on flood frequency: Comparative multisite analysis of dominant runoff processes

We present results of a comparative modeling analysis of the effects of multiscale rainfall variability (within‐event, between‐event, seasonal, interannual, and interdecadal) on estimated flood frequency curves for three catchments located in Perth, Newcastle, and Darwin, Australia. The analysis is performed using the derived distribution approach by combining long‐term rainfall time series generated by a stochastic rainfall model with a continuous rainfall‐runoff flood model that is able to generate runoff variability over a multiplicity of timescales. Similarities and differences of the flood frequency curves (FFCs) in these rather diverse catchments are then interpreted on the basis of differences in the dominant runoff generation processes. In Newcastle, annual maximum flood peaks are caused by saturation excess overland flow over the entire range of annual exceedance probabilities (AEPs) or return periods. On the other hand, in Darwin, the shape of the FFC is determined strongly by seasonal climatic variability, which, in combination with deep soils, leads to a switch of dominant runoff mechanisms contributing to annual maximum flood peaks, from subsurface stormflow at high AEPs (low return periods) to saturation excess overland flow at low AEPs (high return periods). This leads to FFCs exhibiting a consistent break in slope in the Darwin catchment but not so in Newcastle. On the other hand, the FFCs in Perth are affected by both seasonality and long‐term climate variability and produce a variety of shapes depending on the relative strengths of these climatic controls. Because of the fact that in Perth and Darwin the shapes of the flood frequency curves depend on a possible switch of the dominant runoff generation mechanisms with increasing return period, uncertainty in hydrological model parameters relating to landscape properties contributes significantly to the uncertainty in the flood frequency curves. This uncertainty is much less pronounced in Newcastle because of the absence of such a switch of runoff generation mechanisms.

Thresholds in the storm response of a lake chain system and the occurrence and magnitude of lake overflows: Implications for flood frequency

The aim of this paper is to illustrate the effects of spatial organization of lake chains and associated storage thresholds upon lake-overflow behaviour, and specifically their impact upon large scale flow connectivity and the flood frequency of lake overflows. The analysis was carried out with the use of a multiple bucket model of the lake chain system, consisting of a network of both lakes and associated catchment areas, which explicitly incorporated within it three storage thresholds: a catchment field capacity threshold that governs catchment subsurface stormflow, a total storage capacity threshold that governs catchment surface runoff, and a lake storage capacity threshold that determines lake overflow. The model is driven by rainfall inputs generated by a stochastic rainfall model that is able to capture rainfall variability at a wide range of time scales. The study is used to gain insights into the process controls of lake-overflow generation, and in particular, to explore the crucial role of factors relating to lake organization, such as the average catchment area to lake area ( A C/ A L) ratio and the distribution of A C/ A L with distance in the downstream direction (increasing or decreasing). The study showed that the average A C/ A L value was the most important factor determining the frequency of occurrence and magnitude of floods from a landscape consisting of lake chains. The larger the average A C/ A L value the more runoff is generated from catchments thus increasing both the occurrence and magnitude of lake overflows. In this case the flood frequency curve reflects that of the catchment area, and lake organization does not play an important role. When A C/ A L is small the landscape is lake dominated, the spatial organization of lakes has a significant impact on lake connectivity, and consequently on flood frequency. One of the aspects of lake organization that may have a significant influence on lake connectivity is the spatial distribution of A C/ A L from upstream to downstream (increasing or decreasing). In a landscape in which A C/ A L increases downstream, lake overflow will occur more frequently relative to a similar landscape (i.e. identical A C/ A L) with a constant value of A C/ A L. When A C/ A L decreases downstream, however, runoff inputs from the upstream parts will trigger lake overflow in the downstream parts, and consequently, full connectivity may be achieved leading to increased flood frequencies.

Influence of dike breaches on flood frequency estimation

Many river floodplains and their assets are protected by dikes. In case of extreme flood events, dikes may breach and floodwater may spill over into the dike hinterland. Depending on the specific situation, e.g. time and location of breach, and the capacity of the hinterland to contain the floodwater, dike breaches may lead to significant reductions of flood peaks downstream of breach locations. However, the influence of dike breaches on flood frequency distributions along rivers has not been systematically analysed. In order to quantify this influence, a dynamic–probabilistic model is developed. This model combines simplified flood process modules in a Monte Carlo framework. The simplifications allow for the simulation of a large number of different scenarios, taking into account the main physical processes. By using a Monte Carlo approach, frequency distributions can be derived from the simulations. In this way, process understanding and the characteristics of the river–dike–floodplain system are included in the derivation of flood frequency statements. The dynamic–probabilistic model is applied to the Lower Rhine in Germany and compared to the usually used flood frequency analysis. For extreme floods, the model simulates significant retention effects due to dike breaches, which lead to significant modifications of the flood frequency curve downstream of breach locations. The resulting probabilistic statements are much more realistic than those of the flood frequency approach, since the dynamic–probabilistic model incorporates an important flood process, i.e. dike breaching, that only occurs when a certain threshold is reached. Beyond this point, the behaviour of the flood frequency curve is dominated by this process.

Thresholds in the storm response of a catchment‐lake system and the occurrence and magnitude of lake overflows: Implications for flood frequency

This model‐based study examines the combined effects of catchment and lake thresholds upon the frequency and magnitude of lake‐overflow events, and their impacts on flood frequency. A dominant control of lake‐overflow events is antecedent storage, which is governed by the climate and by the properties of the contributing catchment. The next major control was shown to be the magnitude of storm depths, and their adequacy to replenish and exceed the lake storage deficit, which are governed by the ratio of catchment to lake areas, AC/AL. When AC/AL is large, lake‐overflows can be triggered even at larger antecedent lake storage deficits. This points to the importance of AC/AL as a critical parameter governing the frequency and magnitude of lake‐overflow events. As regards the catchment properties, model simulations indicate that fast draining catchments enhance the triggering of lake‐overflow events due to the fact that the drainage is rapid and there is the opportunity for faster runoff contributions to combine with direct rainfall on the lakes to exceed antecedent lake storage deficit and overcome the reducing influence of evaporation during inter‐storm periods. Slow draining catchments, on the other hand, will not release runoff fast enough to replenish the lake storage deficit before the evaporative effects take hold. The shape of the resulting flood frequency curve captures the dominant lake‐overflow generating mechanisms. When the dominant lake overflow generating mechanism is catchment runoff, the shape of flood frequency curve of lake‐overflows resembles the shape of the catchment flood frequency curve, including the effects of associated thresholds. On the other hand, when the dominant lake overflow generating mechanism is direct rainfall falling on the lake, the shape of the lake‐overflow flood frequency curve exhibits a persistent truncation below a critical return period that is associated with the frequency of flow termination for the given climate in question. These results have provided valuable insights into the relative roles of climate, soil depth, the soil's drainage capacity as well as the ratio of catchment area to lake area on flood frequency of catchment‐lake systems in general. The improved understanding of these process controls will be useful in assisting the management of such combined catchment‐lake systems. In particular, they can provide valuable guidance towards the monitoring of catchment‐lake systems in ways that are targeted towards those controls which are critical to the determination of the magnitude and frequency of lake‐overflow events to assist in flood prevention and mitigation.

Derivation of flood frequency curves in poorly gauged Mediterranean catchments using a simple stochastic hydrological rainfall-runoff model

In this paper a Monte Carlo procedure for deriving frequency distributions of peak flows using a semi-distributed stochastic rainfall-runoff model is presented. The rainfall-runoff model here used is very simple one, with a limited number of parameters and practically does not require any calibration, resulting in a robust tool for those catchments which are partially or poorly gauged. The procedure is based on three modules: a stochastic rainfall generator module, a hydrologic loss module and a flood routing module. In the rainfall generator module the rainfall storm, i.e. the maximum rainfall depth for a fixed duration, is assumed to follow the two components extreme value (TCEV) distribution whose parameters have been estimated at regional scale for Sicily. The catchment response has been modelled by using the Soil Conservation Service-Curve Number (SCS-CN) method, in a semi-distributed form, for the transformation of total rainfall to effective rainfall and simple form of IUH for the flood routing. Here, SCS-CN method is implemented in probabilistic form with respect to prior-to-storm conditions, allowing to relax the classical iso-frequency assumption between rainfall and peak flow. The procedure is tested on six practical case studies where synthetic FFC (flood frequency curve) were obtained starting from model variables distributions by simulating 5000 flood events combining 5000 values of total rainfall depth for the storm duration and AMC (antecedent moisture conditions) conditions. The application of this procedure showed how Monte Carlo simulation technique can reproduce the observed flood frequency curves with reasonable accuracy over a wide range of return periods using a simple and parsimonious approach, limited data input and without any calibration of the rainfall-runoff model.

Nonparametric Monte Carlo Simulation for Flood Frequency Curve Derivation: An Application to a Korean Watershed1

Abstract: A mix of causative mechanisms may be responsible for flood at a site. Floods may be caused because of extreme rainfall or rain on other rainfall events. The statistical attributes of these events differ according to the watershed characteristics and the causes. Traditional methods of flood frequency analysis are only adequate for specific situations. Also, to address the uncertainty of flood frequency estimates for hydraulic structures, a series of probabilistic analyses of rainfall‐runoff and flow routing models, and their associated inputs, are used. This is a complex problem in that the probability distributions of multiple independent and derived random variables need to be estimated to evaluate the probability of floods. Therefore, the objectives of this study were to develop a flood frequency curve derivation method driven by multiple random variables and to develop a tool that can consider the uncertainties of design floods. This study focuses on developing a flood frequency curve based on nonparametric statistical methods for the estimation of probabilities of rare floods that are more appropriate in Korea. To derive the frequency curve, rainfall generation using the nonparametric kernel density estimation approach is proposed. Many flood events are simulated by nonparametric Monte Carlo simulations coupled with the center Latin hypercube sampling method to estimate the associated uncertainty. This study applies the methods described to a Korean watershed. The results provide higher physical appropriateness and reasonable estimates of design flood.

Threshold effects in catchment storm response and the occurrence and magnitude of flood events: implications for flood frequency

Abstract. The aim of this paper is to illustrate the effects of selected catchment storage thresholds upon runoff behaviour, and specifically their impact upon flood frequency. The analysis is carried out with the use of a stochastic rainfall model, incorporating rainfall variability at intra-event, inter-event and seasonal timescales, as well as infrequent summer tropical cyclones, coupled with deterministic rainfall-runoff models that incorporate runoff generation by both saturation excess and subsurface stormflow mechanisms. Changing runoff generation mechanisms (i.e. from subsurface flow to surface runoff) associated with a given threshold (i.e. saturation storage capacity) is shown to be manifested in the flood frequency curve as a break in slope. It is observed that the inclusion of infrequent summer storm events increases the temporal frequency occurrence and magnitude of surface runoff events, in this way contributing to steeper flood frequency curves, and an additional break in the slope of the flood frequency curve. The results of this study highlight the importance of thresholds on flood frequency, and provide insights into the complex interactions between rainfall variability and threshold nonlinearities in the rainfall-runoff process, which are shown to have a significant impact on the resulting flood frequency curves.

A conceptual investigation of process controls upon flood frequency: role of thresholds

Abstract. Traditional statistical approaches to flood frequency inherently assume homogeneity and stationarity in the flood generation process. This study illustrates the impact of heterogeneity associated with threshold non-linearities in the storage-discharge relationship associated with the rainfall-runoff process upon flood frequency behaviour. For a simplified, non-threshold (i.e. homogeneous) scenario, flood frequency can be characterised in terms of rainfall frequency, the characteristic response time of the catchment, and storm intermittency, modified by the relative strength of evaporation. The flood frequency curve is then a consistent transformation of the rainfall frequency curve, and could be readily described by traditional statistical methods. The introduction of storage thresholds, namely a field capacity storage and a catchment storage capacity, however, results in different flood frequency "regions" associated with distinctly different rainfall-runoff response behaviour and different process controls. The return period associated with the transition between these regions is directly related to the frequency of threshold exceedence. Where threshold exceedence is relatively rare, statistical extrapolation of flood frequency on the basis of short historical flood records risks ignoring this heterogeneity, and therefore significantly underestimating the magnitude of extreme flood peaks.

Use of a grid‐based hydrological model and regional climate model outputs to assess changing flood risk

AbstractA grid‐based flow routing and runoff‐production model, configured to employ regional climate model (RCM) precipitation estimates as input, is used to assess the effects of climate change on river flows in catchments across the UK. This model, the Grid‐to‐Grid model or G2G, has previously been calibrated and assessed with respect to observed river flows under current climate conditions. Here, the G2G distributed model together with a lumped catchment model, the parameter‐generalized PDM, are applied to simulate river flow and derive flood frequency curves. Two sets of RCM precipitation time‐series, on a 25‐km grid and at hourly intervals, are used: (1) Current (1961–1990) and (2) Future (2071–2100). The effect of one extreme rainfall event in the current precipitation series is to raise the estimated peak flows for some catchments for high return periods under present day rainfall conditions. The future flow series does not contain a comparable flow peak. This significantly affects comparison of the flood frequency curves derived from the flow simulations obtained using the current and future precipitation estimates. Such variability in the results and the dependence on one or two extreme rainfall events emphasizes the need to examine more than one set of current/future precipitation scenarios for flood impact studies. In the absence of a formal ensemble of climate predictions, a resampling method is used to investigate the robustness of the modelled changes in flood frequency. Changes in flood frequency at higher return periods are, not surprisingly, found to be generally less robust than at lower return periods. This is particularly the case for catchments in the south and east of England, which were especially affected by the extreme rainfall event in the current precipitation series. Copyright © 2007 Royal Meteorological Society