scholarly journals A Method for Estimating the Regional Initial and Constant Loss for Design Flood Estimation in West Peninsular Malaysia

2013 ◽  
Vol 74 (2) ◽  
Author(s):  
Sazali Osman ◽  
Ismail Abustan ◽  
Rozi Abdullah
Keyword(s):  
Peninsular Malaysia ◽  
Flood Frequency ◽  
Percentage Error ◽  
Rainfall Runoff ◽  
Design Flood ◽  
Frequency Method ◽  
Flood Estimation ◽  
Common Technique ◽  
Loss Method ◽  
Regional Design

Floods are known as one of the world’s most frequent and devastating events. Techniques to predict and estimate the size of floods is depend on the availability of hydrological data. Using the conceptual of lump model, rainfall-runoff method is widely used in design flood estimation, which represents the input of rainfall and catchment characteristics such as rainfall depth, rainfall intensity, baseflow and losses. 7o calculate the catchments runoff, amount of losses shall be determine accurately by considering various source of the rainfall losses such as evaporation, inßltration, interception, depression storage and loss in groundwater recharge. In Malaysia, the common technique to estimate the hydrological losses is using initial and constant loss method. Furthermore, the value has been used in Urban Stormwater Manual for Malaysia (USMA) are adopted from the other literatures which is not represented the value from local catchment.7he objective of this study is to derive the initial and constant loss values using the data from selected local catchments in west Peninsular Malaysia. 7he calculated initial and constant loss will be further used to derive design flood discharge based on the design rainfall. An initial loss and constant loss model was examined in this study to observe the loss rate parameters in heterogeneous catchments and evaluate their signißcance as well as their potential influence on design peak floods. 7he study has been utilised the rainfall and runoff data from 113 storms over 15 catchments. 7he loss parameters were obtained from model optimi3ation using the HEC-HMS Modeling program. From the analyses, the median initial loss is 21.54mm with the standard deviation 7.85mm. 7he value shows higher than the value adopted in USMA. Meanwhile, the value for constant loss is 8.07mm which between the range of USMA. Based on the ßndings of design initial loss analyses, the values of initial loss were 49.3, 57.6, 64.1, 69.4, 73.3 and 76.6 for ARI 2,5,10,20,50 and 100, respectively. 7he percentage error between design initial loss and constant loss method and flood frequency method shows good results which are most of the percentage error less than 35%. It shows that the design initial loss and constant loss method produce reasonable accurate results when compared to the rainfall-runoff method and flood frequency method. Based on the ßndings, it can be suggested that the regional design initial loss and the constant loss rates would be able to serve reasonably well in determining catchment loss for the design purposes.

Runoff models and flood frequency statistics for design flood estimation in Austria – Do they tell a consistent story?

2012 ◽  
Vol 456-457 ◽  
pp. 30-43 ◽  
Author(s):  
M. Rogger ◽  
B. Kohl ◽  
H. Pirkl ◽  
A. Viglione ◽  
J. Komma ◽  
...  
Keyword(s):  
Flood Frequency ◽  
Design Flood ◽  
Flood Estimation

Comparing runoff statistics simulated with a hydrological model utilizing stochastically generated lumped catchment rainfall and multiple point rainfall

2020 ◽  
Author(s):  
Luisa-Bianca Thiele ◽  
Ross Pidoto ◽  
Uwe Haberlandt
Keyword(s):  
Time Series ◽  
Hydrological Modelling ◽  
Flood Frequency ◽  
Multiple Point ◽  
Rainfall Time Series ◽  
Rainfall Runoff ◽  
Design Flood ◽  
Catchment Scale ◽  
Time Step ◽  
Rainfall Runoff Model

<p>For derived flood frequency analyses, stochastic rainfall models can be linked with rainfall-runoff models to improve the accuracy of design flood estimations when the length of observed rainfall and runoff data is not sufficient. In the past, when using stochastic rainfall time series for hydrological modelling purposes, catchment rainfall for use in hydrological modelling was calculated from the multiple point rainfall time series. As an alternative to this approach, it will be tested whether catchment rainfall can be modelled directly, negating the drawbacks (and need) encountered in generating spatially consistent time series. An Alternating Renewal rainfall model (ARM) will be used to generate multiple point and lumped catchment rainfall time series in hourly resolution. The generated rainfall time series will be used to drive the rainfall-runoff model HBV-IWW with an hourly time step for mesoscale catchments in Germany. Validation will be performed by comparing modelled runoff regarding runoff and flood statistics using stochastically generated lumped catchment rainfall versus multiple point rainfall. It would be advantageous if the results based on catchment rainfall are comparable to those using multiple point rainfall, so catchment rainfall could be generated directly with the stochastic rainfall models. Extremes at the catchment scale may also be better represented if catchment rainfall is generated directly.</p>

scholarly journals Use of historical data in flood frequency analysis: a case study for four catchments in Norway

2017 ◽  
Vol 49 (2) ◽  
pp. 466-486 ◽  
Author(s):  
Kolbjørn Engeland ◽  
Donna Wilson ◽  
Péter Borsányi ◽  
Lars Roald ◽  
Erik Holmqvist
Keyword(s):  
Flood Frequency ◽  
Water Levels ◽  
Added Value ◽  
Flood Frequency Analysis ◽  
Return Periods ◽  
Design Flood ◽  
Historical Information ◽  
Improve Design ◽  
Estimation Uncertainty ◽  
Flood Estimation

Abstract There is a need to estimate design floods for areal planning and the design of important infrastructure. A major challenge is the mismatch between the length of the flood records and needed return periods. A majority of flood time series are shorter than 50 years, and the required return periods might be 200, 500, or 1,000 years. Consequently, the estimation uncertainty is large. In this paper, we investigated how the use of historical information might improve design flood estimation. We used annual maximum data from four selected Norwegian catchments, and historical flood information to provide an indication of water levels for the largest floods in the last two to three hundred years. We assessed the added value of using historical information and demonstrated that both reliability and stability improves, especially for short record lengths and long return periods. In this study, we used information on water levels, which showed the stability of river profiles to be a major challenge.

scholarly journals Design flood estimation with varying record lengths in Norway under stationarity and nonstationarity scenarios

2021 ◽  
Author(s):  
Lei Yan ◽  
Lihua Xiong ◽  
Gusong Ruan ◽  
Chong-Yu Xu ◽  
Mengjie Zhang
Keyword(s):  
Frequency Analysis ◽  
Location Parameter ◽  
Extreme Value ◽  
Flood Frequency ◽  
Flood Frequency Analysis ◽  
Design Flood ◽  
Generalized Extreme Value ◽  
Record Length ◽  
Flood Estimation ◽  
The Stability

Abstract In traditional flood frequency analysis, a minimum of 30 observations is required to guarantee the accuracy of design results with an allowable uncertainty; however, there has not been a recommendation for the requirement on the length of data in NFFA (nonstationary flood frequency analysis). Therefore, this study has been carried out with three aims: (i) to evaluate the predictive capabilities of nonstationary (NS) and stationary (ST) models with varying flood record lengths; (ii) to examine the impacts of flood record lengths on the NS and ST design floods and associated uncertainties; and (iii) to recommend the probable requirements of flood record length in NFFA. To achieve these objectives, 20 stations with record length longer than 100 years in Norway were selected and investigated by using both GEV (generalized extreme value)-ST and GEV-NS models with linearly varying location parameter (denoted by GEV-NS0). The results indicate that the fitting quality and predictive capabilities of GEV-NS0 outperform those of GEV-ST models when record length is approximately larger than 60 years for most stations, and the stability of the GEV-ST and GEV-NS0 is improved as record lengths increase. Therefore, a minimum of 60 years of flood observations is recommended for NFFA for the selected basins in Norway.

scholarly journals Design flood estimation for global river networks based on machine learning models

2020 ◽  
Author(s):  
Gang Zhao ◽  
Paul Bates ◽  
Jeffrey Neal ◽  
Bo Pang
Keyword(s):  
Machine Learning ◽  
Regression Models ◽  
Flood Hazard ◽  
Model Development ◽  
Flood Frequency ◽  
Support Vector ◽  
Design Flood ◽  
Flood Estimation ◽  
Design Floods ◽  
Anderson Darling

Abstract. Design flood estimation is a fundamental task in hydrology. In this research, we propose a machine learning based approach to estimate design floods globally. This approach mainly involves three stages: (i) estimating at-site flood frequency curve for global gauging stations by the Anderson-Darling test and Bayesian MCMC method; (ii) clustering these stations into subgroups by a K-means model based on twelve globally available catchment descriptors, and (iii) developing a regression model in each subgroup for regional design flood estimation using the same descriptors. A total of 11793 stations globally were selected for model development and three widely used regression models were compared for design flood estimation. The results showed that: (1) the proposed approach achieved the highest accuracy for design flood estimation when using all twelve descriptors for clustering; and the performance of regression was improved by considering more descriptors during the training and validation; (2) a support vector machine regression provide the highest prediction performance among all regression models tested, with root mean square normalised error of 0.708 for 100-year return period flood estimation; (3) 100-year design flood in tropical, arid, temperate, cold and polar climate zones could be reliably estimated with the relative mean relative biases (RBIAS) of −0.199, −0.233, −0.169, 0.179 and −0.091 respectively; (4) This machine learning based approach shows considerable improvement over the index-flood based method introduced by Smith et al. (2015, https://doi.org/10.1002/2014WR015814) for the design flood estimation at global scales; and the average RBIAS in estimation is less than 18 % for 10, 20, 50 and 100-year design floods. We conclude that the proposed approach is a valid method to estimate design floods anywhere on the global river network, improving our prediction of the flood hazard, especially in ungauged areas.

scholarly journals Effects of Probability-Distributed Losses on Flood Estimates Using Event-Based Rainfall-Runoff Models

Water ◽  
2021 ◽  
Vol 13 (15) ◽  
pp. 2049
Author(s):  
Melanie Loveridge ◽  
Ataur Rahman
Keyword(s):  
Probability Distributions ◽  
Data Sets ◽  
Rainfall Runoff ◽  
Design Flood ◽  
Ungauged Catchments ◽  
Large Dataset ◽  
Loss Data ◽  
Flood Estimation ◽  
Historic Data ◽  
Event Based

Probability distributions of initial losses are investigated using a large dataset of catchments throughout Australia. The variability in design flood estimates caused by probability-distributed initial losses and associated uncertainties are investigated. Based on historic data sets in Australia, the Gamma and Beta distributions are found to be suitable for describing initial loss data. It has also been found that the central tendency of probability-distributed initial loss is more important in design flood estimation than the form of the probability density function. Findings from this study have notable implications on the regionalization of initial loss data, which is required for the application of Monte Carlo methods for design flood estimation in ungauged catchments.

scholarly journals Climate, orography and scale controls on flood frequency in Triveneto (Italy)

2016 ◽  
Vol 373 ◽  
pp. 95-100 ◽  
Author(s):  
Simone Persiano ◽  
Attilio Castellarin ◽  
Jose Luis Salinas ◽  
Alessio Domeneghetti ◽  
Armando Brath
Keyword(s):  
Flood Frequency ◽  
Homogeneous Region ◽  
Annual Maximum ◽  
Design Flood ◽  
Reference Procedure ◽  
North Eastern ◽  
Frequency Regime ◽  
Flood Estimation ◽  
L Moments ◽  
Annual Maximum Series

Abstract. The growing concern about the possible effects of climate change on flood frequency regime is leading Authorities to review previously proposed reference procedures for design-flood estimation, such as national flood frequency models. Our study focuses on Triveneto, a broad geographical region in North-eastern Italy. A reference procedure for design flood estimation in Triveneto is available from the Italian NCR research project "VA.PI.", which considered Triveneto as a single homogeneous region and developed a regional model using annual maximum series (AMS) of peak discharges that were collected up to the 1980s by the former Italian Hydrometeorological Service. We consider a very detailed AMS database that we recently compiled for 76 catchments located in Triveneto. All 76 study catchments are characterized in terms of several geomorphologic and climatic descriptors. The objective of our study is threefold: (1) to inspect climatic and scale controls on flood frequency regime; (2) to verify the possible presence of changes in flood frequency regime by looking at changes in time of regional L-moments of annual maximum floods; (3) to develop an updated reference procedure for design flood estimation in Triveneto by using a focused-pooling approach (i.e. Region of Influence, RoI). Our study leads to the following conclusions: (1) climatic and scale controls on flood frequency regime in Triveneto are similar to the controls that were recently found in Europe; (2) a single year characterized by extreme floods can have a remarkable influence on regional flood frequency models and analyses for detecting possible changes in flood frequency regime; (3) no significant change was detected in the flood frequency regime, yet an update of the existing reference procedure for design flood estimation is highly recommended and we propose the RoI approach for properly representing climate and scale controls on flood frequency in Triveneto, which cannot be regarded as a single homogeneous region.

scholarly journals Design flood estimation for global river networks based on machine learning models

2021 ◽  
Vol 25 (11) ◽  
pp. 5981-5999
Author(s):  
Gang Zhao ◽  
Paul Bates ◽  
Jeffrey Neal ◽  
Bo Pang
Keyword(s):  
Machine Learning ◽  
Regression Models ◽  
Flood Hazard ◽  
Model Development ◽  
Flood Frequency ◽  
Support Vector ◽  
Design Flood ◽  
Flood Estimation ◽  
Design Floods ◽  
Anderson Darling

Abstract. Design flood estimation is a fundamental task in hydrology. In this research, we propose a machine-learning-based approach to estimate design floods globally. This approach involves three stages: (i) estimating at-site flood frequency curves for global gauging stations using the Anderson–Darling test and a Bayesian Markov chain Monte Carlo (MCMC) method; (ii) clustering these stations into subgroups using a K-means model based on 12 globally available catchment descriptors; and (iii) developing a regression model in each subgroup for regional design flood estimation using the same descriptors. A total of 11 793 stations globally were selected for model development, and three widely used regression models were compared for design flood estimation. The results showed that (1) the proposed approach achieved the highest accuracy for design flood estimation when using all 12 descriptors for clustering; and the performance of the regression was improved by considering more descriptors during training and validation; (2) a support vector machine regression provided the highest prediction performance amongst all regression models tested, with a root mean square normalised error of 0.708 for 100-year return period flood estimation; (3) 100-year design floods in tropical, arid, temperate, cold and polar climate zones could be reliably estimated (i.e. <±25 % error), with relative mean bias (RBIAS) values of −0.199, −0.233, −0.169, 0.179 and −0.091 respectively; (4) the machine-learning-based approach developed in this paper showed considerable improvement over the index-flood-based method introduced by Smith et al. (2015, https://doi.org/10.1002/2014WR015814) for design flood estimation at global scales; and (5) the average RBIAS in estimation is less than 18 % for 10-, 20-, 50- and 100-year design floods. We conclude that the proposed approach is a valid method to estimate design floods anywhere on the global river network, improving our prediction of the flood hazard, especially in ungauged areas.

Exploring Uncertainties in Flood Frequency Analysis Using Bayesian MCMC Method

2013 ◽  
Vol 663 ◽  
pp. 768-772
Author(s):  
Li Jie Zhang
Keyword(s):  
Frequency Analysis ◽  
Extreme Value Distribution ◽  
Flood Frequency ◽  
Generalized Extreme Value Distribution ◽  
Flood Frequency Analysis ◽  
Distribution Model ◽  
Mcmc Method ◽  
Design Flood ◽  
Pearson Type ◽  
Flood Estimation

The evaluation and reducing of uncertainty is central to the task of hydrological frequency analysis. In this paper a Bayesian Markov Chain Monte Carlo (MCMC) method is employed to infer the parameter values of the probabilistic distribution model and evalue the uncertainties of design flood. Comparison to the estimated results of three-parameter log-normal distribution (LN3) and the three-parameter generalized extreme value distribution (GEV), the Pearson Type 3 distribution (PIII) provides a good approximation to flood-flow data. The choice of the appropriate probabilistic model can reduce uncertainty of design flood estimation. Historical flood events might be greatly reduced uncertainty when incorporating past extreme historical data into the flood frequency analysis.

scholarly journals A systematic comparison of statistical and hydrological methods for design flood estimation

2019 ◽  
Vol 50 (6) ◽  
pp. 1665-1678 ◽  
Author(s):  
Kenechukwu Okoli ◽  
Maurizio Mazzoleni ◽  
Korbinian Breinl ◽  
Giuliano Di Baldassarre
Keyword(s):  
Model Selection ◽  
Statistical Methods ◽  
Model Averaging ◽  
Return Periods ◽  
Rainfall Runoff ◽  
Design Flood ◽  
High Return ◽  
Flood Estimation ◽  
Rainfall Runoff Model ◽  
Runoff Model

Abstract We compare statistical and hydrological methods to estimate design floods by proposing a framework that is based on assuming a synthetic scenario considered as ‘truth’ and use it as a benchmark for analysing results. To illustrate the framework, we used probability model selection and model averaging as statistical methods, while continuous simulations made with a simple and relatively complex rainfall–runoff model are used as hydrological methods. The results of our numerical exercise show that design floods estimated by using a simple rainfall–runoff model have small parameter uncertainty and limited errors, even for high return periods. Statistical methods perform better than the linear reservoir model in terms of median errors for high return periods, but their uncertainty (i.e., variance of the error) is larger. Moreover, selecting the best fitting probability distribution is associated with numerous outliers. On the contrary, using multiple probability distributions, regardless of their capability in fitting the data, leads to significantly fewer outliers, while keeping a similar accuracy. Thus, we find that, among the statistical methods, model averaging is a better option than model selection. Our results also show the relevance of the precautionary principle in design flood estimation, and thus help develop general recommendations for practitioners and experts involved in flood risk reduction.

Sign in / Sign up

Export Citation Format

Share Document