Flood frequency analysis using mean daily flows vs. instantaneous peak flows

<p>Flood frequency analysis (FFA) has long been the standard procedure for obtaining design floods for all kinds of purposes. Ideally, the data at the basis of the statistical operations have a high temporal resolution, in order to facilitate a full account of the observed flood peaks and hence a precise model fitting and flood quantile estimation.</p><p>Unfortunately, high-resolution flows are rarely disposable. Often, average daily flows pose the only available/sufficiently long base for flood frequency analysis. This averaging naturally causes a significant smoothing of the flood wave, such that the “instantaneous” peak can no longer be observed. As a possible consequence, design floods derived from these data may be severely underrated.</p><p>How strongly the original peaks are flattened and how this influences the design flood estimation depends on a variety of factors and varies from gauge to gauge. In this study we are looking at a range of errors arising from the use of daily instead of instantaneous flow data. These include differences in the observed individual flood peaks and mean annual maximum floods, as well as the estimated distribution parameters and flood quantiles. The aim is to identify catchment specific factors that influence the magnitude of these errors, and ultimately to provide a means for error assessment on the mere basis of local hydrological conditions, specifically where no high-resolution data is available.</p><p>The analyses are carried out on an all-German dataset of discharge gauges, for which high-resolution data is available for at least 30 years. The classical FFA approach of fitting distributions to annual maximum series is utilized for error assessment. For identification of influencing factors, both the discharge series themselves and a catalogue of climatic and physiographic catchment descriptors are screened.</p>

Estimates of annual maximum (peak) flow quantiles are needed for basins undergoing changes in both urbanization and climate. Most previous work on the effect of urbanization on peak flows has considered urbanization alone and only the spatial variation in flood quantiles or its mean temporal effect, and most work on the effect of nonstationarity in climate has focused on single-station analyses, which give uncertain results for extreme quantiles. To address these gaps, three approaches to the statistical estimation of the joint effects of changes in impervious cover and climate on the estimation of peak-flow quantiles were compared: single-station quantile regression; a fixed effect panel-quantile regression (pQR) method using a location (mean) shift to homogenize the panel; and a location-scale panel regression model (pQRmom), which accounts for both scale (variance) and location effects. The different approaches were applied to a dataset consisting of instantaneous annual peak flows from 127 minimally nested basins in the midwestern United States with at least 4 % change in imperviousness. The annual maximum daily discharge from a water-balance model was selected as the primary climate predictor; in addition, to provide a comparison of climate predictors, precipitation was also considered. The coefficients from single-station regressions were usually sufficiently certain to determine the effects of climate variation but usually too uncertain to estimate the effects of urbanization. The panel-quantile regression approaches give much more certain results, but their estimates of quantile dependence differ: although both indicate urbanization effects decreasing with decreasing annual exceedance probability (AEP), the pQRmom urbanization coefficients are insignificantly different from zero for AEPs less than 0.10, whereas the pQR coefficients remain positive and are significant except for AEP = 0.01, the smallest AEP value considered. Although the location-scale structure of the pQRmom approach has less flexible quantile dependence than the pQR approach, the pQRmom approach has somewhat lower overall error, and it is found that by subsetting the dataset to homogenize the scale effects, the pQR and pQRmom results become similar, indicating the insignificant urbanization coefficients for small AEPs of the pQRmom results are likely correct for the study dataset.

Analyses of the magnitude and frequency of a 400-year flood record in the Fish River Basin, Namibia

<p>In the last few years, several studies have detected changes in the rainfall regime which may have an impact on hydrological extremes and water resources availability. In addition, the lack of continuous observations as well as the significant transformations of river basins limits our ability to fully characterize hydrological response. Therefore, it is urgent to update current methods and tools in order to shed light on the expected hydrological changes.</p> <p>In the present study, we try to construct a detailed description of extreme flow patterns in Southern Italy in the period 1920-2020. For this reason, the dataset of annual maximum discharges was constructed using all available records and extended using indirect measurements (e.g., daily discharge and water levels). The data before 1980 were collected in the SIMN Special Publication No. 17 and in the SIMN Hydrological Yearbooks, which provide the annual maximum flow rates and the annual rating curves. Hydrological observatory have been transferred to the regional Department of Civil Protection, but only water level observations are available for the most recent period.</p> <p>Dataset from different sources requires a significant effort to reconstruct reliable timeseries. In order to extend the historical series of floods annual maxima, we tried to transform mean daily maxima into peak flow values by means of a conversion factor proposed by Taguas et al. (2008). Additionally, the database was also integrated with the most recent data converting water level measurements in annual floods by using the annual maxima flow rating curve. Such rating curve turned out to be quite stable over time as demonstrated by Claps et al. (2010) and it was verified using also hydraulic numerical models.</p> <p>The present study results complement the outcomes of the recent study by Blöschl et al. (2017), who investigated flood trends over the last five decades in Europe. This study provided a clear overview on the recent tendencies in Europe except for southern Italy because of the limited and discontinuous data availability. Therefore, the study allowed to reconstruct a relevant number of timeseries representative of the entire southern Italy. The homogeneity of the reconstructed data have been verified using the Kolmogorov-Smirnov test. Then, the obtained series were analysed in order to detect possible trends by using the Mann-Kendall non-parametric test. Results highlights the dynamics of flood production over the entire southern Italy.</p> <p> </p> <p><strong>References</strong></p> <p>Blöschl G., Hall J., Parajka J., Perdigão R.A.P., Merz B, Arheimer B., Aronica G.T., Bilibashi A., Bonacci O., [...], Živković N. (2017). Changing climate shifts timing of European floods. Science, 357, pp. 588-59.</p> <p>Claps P., Ganora D., Laio F., and Radice R. (2010) Riesame ed integrazione di serie di portate al colmo mediante scale di deflusso di piena, Atti del XXXII Convegno Nazionale di Idraulica e Costruzioni Idrauliche, Idraulica e Costruzioni Idrauliche, Palermo, 14-17 settembre 2010.</p> <p>Taguas E.V., Ayuso J.L., Pena A., Yuan Y., Sanchez M.C., Giraldez J.V., and Pérez R. (2008) Testing the relationship between instantaneous peak flow and mean daily flow in a Mediterranean Area Southeast Spain, Catena, 75, pp. 129-137. https://doi.org/10.1016/j.catena.2008.04.015</p>

Flood trends are commonly assessed based on instantaneous peak flows on an hourly timescale, as these are most relevant for flood management. However, when hourly data are missing, it has been suggested to perform flood statistics on daily flood values instead, assuming a scaling relationship that depends on the shape of the flood hydrograph and applies over the entire observation period (e.g. Bartens & Haberlandt, 2021). In an Austria-wide assessment, recent flood trends show diverging spatial patterns that contradict such a stationarity assumption. Interestingly, an aggravation of the flood situation is mainly observed for the peak flow (IPF), while the high values of the mean daily discharge (MDF) show much smaller and, importantly, less significant trends. Rather than applying flood statistics corrections (e.g. Beylich et al. 2021), the aim of this contribution is to use flood divergence at different timescales as a mean of inferring likely drivers of flood trends. To this end, we combine several established and innovative indicators, such as a trend divergence index (peak versus daily flood scale), a seasonal trend index (to infer information about flood generation processes), and a seasonal shift index (to infer changes in the relevance of these processes). We show the extent to which these indices can inform us about likely drivers of change, i.e. climate-related vs. anthropogenic changes in the catchment. We discuss how these indicators perform in the light of existing flood scale indices, such as the flood timescale (Gaál et al., 2012) and the peak-volume ratio (Bartens & Haberlandt, 2021). The results suggest that the conflicting space-time patterns contain a wealth of information that is highly informative about changes in flood controls under global change.

Flood frequency analysis estimates reliable design flood, which is critical for the safety of any hydraulic, hydrologic construction and management. The overall objective of this study is to perform flood frequency analysis for Seonath and Hasdeo river basins of Chhattisgarh state, India. In this study, homogeneous regions are delineated by Ward cluster analysis. Cluster variables are L-moment ratios of annual maximum events and catchment characters. Index flood is performed for parametric quantile estimation of flood events. L-moments are employed in regionalization, frequency distribution selection, and distribution parameter estimation. Locally weighted polynomial model (Locfit) is performed for at-site frequency analysis. Flood quantiles are estimated for regions and at sites. In regional flood frequency analysis, regionalisation based on annual maximum flood series shows better flood quantiles than regionalisation based on site characteristics. It reveals that geophysical homogeneity does not mean hydrologic homogeneity. Regional formulae for annual mean discharge are formulated based on catchment character for estimation of flood quantiles for ungauged catchments.KeywordsFlood frequency analysisCluster analysisL-momentsIndex floodLocfit

First posted January 25, 2024 For additional information, contact: Director, Dakota Water Science CenterU.S. Geological Survey821 East Interstate Avenue, Bismarck, ND 585031608 Mountain View Road, Rapid City, SD 57702Contact Pubs Warehouse Flood-frequency analysis, also called peak-flow frequency or flood-flow frequency analysis, is essential to water resources management applications including critical structure design and floodplain mapping. Federal guidelines for doing flood-frequency analyses are presented in a U.S. Geological Survey Techniques and Methods Report known as Bulletin 17C. A basic assumption within Bulletin 17C is that for drainage basins without major hydrologic alterations, statistical properties of the distribution of annual peak streamflows (peak flows) are stationary; that is, the mean, variance, and skew are constant. The stationarity assumption has been widely accepted within the flood-frequency community; however, a better understanding of long-term climatic persistence and concerns about potential climate change and land-use change has caused a reexamination of the stationarity assumption. Flood-frequency analyses that do not incorporate observed trends and abrupt changes may result in a poor representation of the true flood risk. Bulletin 17C does not offer guidance for incorporating nonstationarities when estimating floods, and it describes a need for studies that incorporate changing climate or basin characteristics. In response to this need and a history of concern regarding nonstationarity peak flows in the region, this study was done by the U.S. Geological Survey, in cooperation with the Departments of Transportation of Illinois, Iowa, Michigan, Minnesota, Missouri, South Dakota, and Wisconsin; the Montana Department of Natural Resources and Conservation; and the North Dakota Department of Water Resources, to assess potential nonstationarity in peak flows in the north-central United States.This chapter summarizes the methods used to detect hydroclimatic changes in peak-flow data in the study region. A wide range of analyses and statistical approaches are applied to document the primary mechanisms controlling floods and characterize temporal changes in hydroclimatic variables and peak flow. Four periods were selected for analysis of peak flow, daily streamflow, and climate data. The periods are (1) a 100-year period, 1921–2020; (2) a 75-year period, 1946–2020; (3) a 50-year period, 1971–2020; and (4) a 30-year period, 1991–2020. The climate data consist of monthly time series estimates of temperature, precipitation, potential evapotranspiration, actual evapotranspiration, snowfall, soil moisture storage, snow water equivalent, and runoff on a 3.1-mile by 3.1-mile grid for the conterminous United States.Statistical and graphical analyses were used to investigate potential changes in hydrology and climate. The starting point for these analyses was the initial data analysis of peak flow described in Bulletin 17C, which includes plotting the peak flow and checking for autocorrelation, monotonic trends, and changes points. Analyses were added to examine additional features in the data. To examine potential causal drivers of changes, the climate data were analyzed graphically and statistically. Results are provided in a U.S. Geological Survey data release. The study limitations are documented for users of the results.

Ce travail présente l'histoire des crues de l'Oubangui depuis 1889 à Bangui, ainsi que leurs impacts sur les personnes et les biens dans le lit majeur urbanisé. Une analyse documentaire, des entretiens avec la population et des observations de terrain ont permis de préciser les dates d'inondation et l'évolution des enjeux dans le lit majeur de l'Oubangui, au regard des crues de 1999 et 2019 (Qj-max de 12180 et 12400 m3/s respectivement). Un traitement statistique des débits (modules et débits journaliers maximums) sur la période 1911-12 - 2020-21 a conduit à caractériser les crues entre périodes humide, moyenne et sèche. Les crues majeures considérées, dites "historiques", ont des débits supérieurs ou égaux à 12000 m3/s. Onze ont été recensées. Selon la loi de GUMBEL, la plus faible période de retour est de 6,5 ans (Qj-max = 12000 m3/s en 1969) et la plus élevée de 53 ans (Qj-max = 16000 m3/s en 1916). Le Qj-max de 12900 m3/s enregistré en 1969 est proche de la valeur décennale (12200 à 12880 m3/s). Les autres périodes vont jusqu'à 22 ans, pour un Qj-max de 14400 m3/s (en 1961). Pour la crue de 1916, les lois GEV, de WEIBULL et Normale donnent des temps de retour de 700 ans pour la première et de plus de 1000 ans pour les deux autres. Les niveaux de risque ont été spatialisés pour trois crues (1916, 1961 et 1999), ce qui ouvre des pistes pour améliorer une gestion du risque d'inondation, encore très sommaire, à Bangui comme dans tout le pays.

Regional flow duration curve estimation in small ungauged catchments using instantaneous flow measurements and a censored data approach

Regression‐based regional flood frequency analysis (RFFA) methods are widely adopted in hydrology. This paper compares two regression‐based RFFA methods using a Bayesian generalized least squares (GLS) modelling framework; the two are quantile regression technique (QRT) and parameter regression technique (PRT). In this study, the QRT focuses on the development of prediction equations for a flood quantile in the range of 2 to 100 years average recurrence intervals (ARI), while the PRT develops prediction equations for the first three moments of the log Pearson Type 3 (LP3) distribution, which are the mean, standard deviation and skew of the logarithms of the annual maximum flows; these regional parameters are then used to fit the LP3 distribution to estimate the desired flood quantiles at a given site. It has been shown that using a method similar to stepwise regression and by employing a number of statistics such as the model error variance, average variance of prediction, Bayesian information criterion and Akaike information criterion, the best set of explanatory variables in the GLS regression can be identified. In this study, a range of statistics and diagnostic plots have been adopted to evaluate the regression models. The method has been applied to 53 catchments in Tasmania, Australia. It has been found that catchment area and design rainfall intensity are the most important explanatory variables in predicting flood quantiles using the QRT. For the PRT, a total of four explanatory variables were adopted for predicting the mean, standard deviation and skew. The developed regression models satisfy the underlying model assumptions quite well; of importance, no outlier sites are detected in the plots of the regression diagnostics of the adopted regression equations. Based on ‘one‐at‐a‐time cross validation’ and a number of evaluation statistics, it has been found that for Tasmania the QRT provides more accurate flood quantile estimates for the higher ARIs while the PRT provides relatively better estimates for the smaller ARIs. The RFFA techniques presented here can easily be adapted to other Australian states and countries to derive more accurate regional flood predictions. Copyright © 2011 John Wiley & Sons, Ltd.

The upper half of one watershed and the lower half of another were deforested and maintained barren from 1965 to 1967. Forest on the remaining halves was cut in 1968. Water yield increases from both watersheds averaged almost 6 inches during the half deforested stage and rose to over 10 inches after complete deforestation. The duration of the flows greatly increased on both watersheds. Other hydrologic effects (instantaneous peak flows, stream temperature, specific conductance, and turbidity) were greater on the lower half deforested watershed. These results show again that forest cutting causes substantial increases in streamflow, but, more important, they provide some landmark values on hydrologic effects of complete deforestation.

The recurring flooding causes loss of life and damage to buildings and other structures, including bridges, sewerage systems, roadways and canals. It also frequently damages power transmission and sometimes power generation, which then has knock-on effects caused by the loss of power. The present study compares the relative performance of flood frequency methods to estimate design flood, using available data of 18 small catchments in the Mahanadi River basin (India). The primary objective of the referred approach was for design flood estimation at gauge sites; however, the main focus is referred to ungauged catchments by an interpolation method. In this regard, three interpolation methods are used: (1) inverse distance weighing method, (2) ordinary kriging and (3) area weighted method. As per the recent trends, flood frequency analysis methods are used specifically for two data types, i.e., at-site analysis and for regionalizing the available data within the homogeneous region. In this study, an attempt has been made to categorize the interpolation properties, where the first two approaches belong to site analysis and the third one uses the regional analysis. In the first approach, the output results in terms of flood quantiles are interpolated for the intermediate results, which is generally termed as direct interpolation of flood quantile, and the second approach uses the linear interpolating or L-moments in flood estimation. The above one refers to the interpolation of L-moments, while flood index is interpolated in the third approach, which is named as ‘flood index procedure.’ In the study, it was observed that the designed flood quantile results were better by using the flood index approach at lower return periods at 2 and 5 years, and the direct interpolation method gave a better estimation for higher return periods. Further, it was found that the difference in prediction error of direct interpolation of flood quantiles and the flood index procedure is negligible.

Flood risk analysis in the United States follows the guidelines recommended in Bulletin 17B, which inherently ties the probability of exceeding a selected peak flow to the historical record. However, insufficient streamflow data, land use change, or climate change can render these recommended techniques impractical, forcing alternative measures. One such alternative is hydrological models which use climate and landscape data to simulate streamflow. Daily time steps or finer, however, are required to extract the necessary peak flow series. Accurately simulating flows at a daily time step can be challenging, and bias in prediction of hydrological extremes can be large. In this analysis, we compare three types of statistical bias‐correction measures for peak flow series data extracted from simulations using the Soil and Water Assessment Tool (SWAT), a widely used continuous, semi‐distributed hydrologic model. Results show that a correction method that only adjusts the mean of the peak flow series was the most robust. This efficacy is likely a result of persistent bias in watershed simulation due to uncertainty in calibration parameters and simplified watershed processes, suggesting that correcting for the mean of the peak flow series may be sufficient when applying bias corrections for flood risk analysis using hydrological models.

This paper presents the development and validation of an artificial intelligence based regional flood frequency analysis (RFFA) model for application in ungauged catchments of eastern Australia. The artificial intelligence based techniques with a flexible model structure and non-linear approach can overcome the limitations of the conventional RFFA models, which are generally based on linear relationship between flood statistics and catchment characteristics. Till now, there have been limited applications of artificial intelligence based techniques to RFFA problems in Australia. This study has developed four artificial intelligence based RFFA models for eastern Australia, using a comprehensive flood database available as a part of Australian Rainfall and Runoff (ARR) revision 'Project 5 Regional flood methods'. These four RFFA models are based on artificial neural network (ANN), genetic algorithm based artificial neural network (GAANN), gene-expression programing (GEP) and co-active neuron fuzzy inference system (CANFIS). A total of 452 catchments from the states of New South Wales, Victoria, Queensland and Tasmania have been considered by this study. These dataset is divided into training and validation sets. Data of 362 catchments (training data set) have been used to train the model and the data from the remaining 90 catchments (validation data set) used to validate the model. The models have been trained/calibrated using the training data set that involved minimisation of the mean squared error between the observed and predicted flood quantiles by the model (being trained) for a given ARI for the training data set. Six average recurrence intervals (ARI) flood quantiles (2, 5, 10, 20, 50 and 100 years) were considered in this study. Four evaluation statistics are adopted to assess the model accuracy: median ratio of the predicted flood quantile (Qpred) and observed flood quantile (Qobs), denoted by Qpred/Qobs ratio, plots of Qobs and Qpred, median relative error and coefficient of efficiency. This is initially done for the training data set and then repeated for the validation data set. The artificial intelligence based RFFA models have been ranked based on their relative performances in relation to the above criteria to identify the best trained/calibrated model. It has been found that none of the four models is superior across all the six ARIs against the adopted criteria. The ANN based RFFA model has a better ranking score in terms of training/calibration; therefore, it is suggested that the ANN-based RFFA model is the best calibrated model among the four artificial intelligence based RFFA models for eastern Australia.

The River Chenab is one of the main western rivers of the Indus River system in Pakistan, which undergoes intensive inundation almost every year during the late monsoon period. The present study performs flood frequency analyses for the river basin as well as simulates different levels of water flow in the system to speculate all kinds of inundation under different scenarios, i.e., to predict flood hazard and flood extended areas. Flood frequency analyses were performed at MARALA Headworks to Khanki Headworks. Data were collected from the Punjab Irrigation Department, Pakistan and from USGS and ASTER GDEM. The peak discharge of MARALA Headworks had been analyzed for 25 years. The preprocessing was performed in HEC Geo-RAS after preprocessing model run in HEC-RAS. After analysis the data were exported in HEC-RAS to ARCMAP to generate a floodplain and inundation map. Our analysis generated the result that different areas would be under water in different return periods. Flood hazards maps for different return periods 10, 20, 50 and 100 years were conducted using annual peaks flow of 35 years from 1980 to 2016. The maximum discharges at up and down stream for different periods were obtained using Gumbel distribution model results which showed that different areas were predicted under water in different return periods and affected areas after five years' return period.