Hydraulic Methods for Catastrophes: Foods, Droughts, Environmental Disasters
The overall theme of the 28th School of Hydraulics was Hydraulic Methodsfor Catastrophes: Floods, Droughts, Environmental Disasters. Most of the presentationswere therefore devoted to a variety of problems that may contribute to increaseour knowledge in the description or prediction of extreme situations that wedeal with in aquatic environment and where hydraulic methods offer important solutions.All the papers presented during the School (after thorough review procedure)are presented in this volume.
Hydraulic methods for catastrophes: floods, droughts, environmental disasters (Preface)
Series: (E-10), 2008, pp.3-4
Measurement techniques for the estimation of cohesive sediment erosion
Series: (E-10), 2008, pp.5-19
This paper provides an overview on measurement techniques related to cohesive sediment erosion. Important hydraulic parameters governing the erosion potential of cohesive sediments are defined and methods for the determination of these parameters in field studies are discussed. An overview over the available in situ technology is given. For this purpose, the in situ instruments are classified in recirculating flumes, straight flow-through flumes, and miscellaneous devices. Hydraulic working principles, advantages and disadvantages of the devices are described. Results of recent comparative studies are summarized.
Site verified contaminant transport model as a mathematical vehicle for prevention of the natural aquifer contamination
Series: (E-10), 2008, pp.21-27
To prevent the natural aquifer contamination, the chosen site verification of the practical mathematical model (equation) of conservative contaminant transport in a groundwater stream was presented. This model includes, except of the advection and dispersion processes, the source (negative) term of reversible sorption which can be described by the well-known non-linear Freundlich adsorption isotherm in relation to statics of this process. In this 2D-mathematical model the numerical solution (using the finite difference method) was used, based on the previously calculated values of the longitudinal and transverse dispersion coefficients (Dx, Dy) as well as the non-linear adsorption parameters (K, N). The calculated maximal values of chloride concentrations based on this mathematical model (with and without adsorption term) were compared with the measured chloride concentrations in the piezometers installed in the chosen natural aquifer for the site verification.
Particle-particle collision for Lagrangian modelling of saltating grain: theoretical background
Series: (E-10), 2008, pp.29-37
The goal of this paper is to describe the method of application of particle-particle collision for Lagrangian modelling of saltating grain in rivers. The model based on the approach proposed for turbulent gas-solid flow and on statistical physics for interparticle collision in gases is presented. This approach relies on determination of the velocity of the considered particles after collision with another particle during the saltation. These collisions, which depend on the particle concentration, the particle size, and particle velocity are briefly discussed. The formulas for collision probability and the method for calculation of velocity after collision and direction of considered particles before collision are proposed.
Tracing of modelled pollution originating from tributaries in the Dobczyce lake working under flood conditions
Series: (E-10), 2008, pp.39-46
This paper shows an application of the water flow velocity maps model that was designed for the Dobczyce retention water body. The model is used to track simulated pollutants that enter the lake from a number of its tributaries. The idea, its use and sample results are shown. The background model is referenced.
Numerical modelling of pollution transport phenomena in the Lake Dobczyce
Series: (E-10), 2008, pp.47-54
This article describes a model designed for tracking the movement of a few kinds of pollutants in a wide range of environmental conditions (water level, total discharge, operational mode of the lake, presence or absence of wind) present in the Dobczyce lake. To solve the case, a finite element approach is used. Model theory and a sample result are both shown.
Determination of the range of active flow zone in one-dimensional flow models
Series: (E-10), 2008, pp.55-62
This paper presents a method to calculate the active flow zone. The method assumes that the active part of the cross-section consists of the area which determines inbank capacity, and of the zone of interaction between the main channel and the floodplains. Using Pasche’s method, the range of this zone has been determined for various vegetative clusters (trees and bushes) typical for river valleys.
Numerical simulations of storm surge disaster due to typoon Maemi in Korea
Series: (E-10), 2008, pp.63-69
Three dimensional numerical simulations were carried out to calculate the storm surge and inundation area due to Typoon Maemi. The typoon landed on the southern coast of Korean Peninsula at 21 hr on 12 September 2003 with a central pressure of 950 hPa. It caused a terrific life damage with more than 130 people missing and dead and the property damage of about 5 billion US dollars. The residential and commercial area facing the Masan Bay located in the southern coast of Korea was heavily flooded and underground facilities suffered from the inundation by the storm surge. The simulated storm surge and the inundation area showed good agreement with field data.
Random-Vortex Method for Free Surface Boundary Problem
Series: (E-10), 2008, pp.71-82
In this paper the random-vortex method is applied to the two-dimensional flow of an incompressible viscous liquid. A formula for the boundary condition on the free surface and a formula for the no-slip and no-through boundary conditions are derived. The flow problem is solved using a fractional step algorithm. In the first step a velocity field is calculated from a vorticity advection equation. In the next steps the field is modified until it satisfies the boundary conditions. The condition on the free surface consists in the generation of a vortex sheet as a function of the free surface curvature and the liquid velocity components. The condition on the solid boundary is fulfilled by defining a vortex sheet on the boundary through a Fredholm equation of the 2nd kind and by calculating an additional potential velocity field. The diffusion of vorticity is determined by the random walk method. It is shown how the method works using as an example a stationary flat flow around a flap gate. The results of the calculations are compared with measurements performed on a flap gate model and good agreement is obtained.
Modelling of river network with widespread floodplain valleys
Series: (E-10), 2008, pp.83-89
One-dimensional models of unsteady flow provide a good representation of flow transformation as long as the conditions of flow are close to the assumptions of 1D motion, i.e., there is one predominant flow direction which coincides with river longitudinal gradient. However, this assumption is too simplistic when a lengthy river section with widespread floodplain valleys is being modelled. When developing a numerical model of river hydrodynamics, correct representation of flow transformation through these water-course elements is the most important, but also the most challenging part. Motion of liquids in floodplains may not be regarded as one-dimensional (particularly in the initial phase of filling the valley with water and in the closing phase, when water returns to the main channel). Representation of these would require 2D models. One of the methods used to represent spatial motion of water on floodplains in 1D unsteady flow models is to split the river cross-section into its active and inactive zones. The fundamental difficulty is to determine the range of active cross-section. Pasche’s method is one of possible techniques applicable for this purpose.
Hydraulic problems during 2001 flood in Gdańsk
Series: (E-10), 2008, pp.91-99
Floods in urban areas cause considerable economic and social losses. Measures for the mitigation of flood consequences are usually limited by existing city infrastructure. The city of Gdańsk is situated within a complicated system of rivers and channels called Gdańsk Water Node (GWN). The critical point of the GWN is the Radunia Channel, 13.5 km long. The main reason of the urban flash flood was intensive precipitation which appeared on 9 July 2001. The paper presents reasons, run and consequences of the flood, hydrological analysis, field measurements, formulation of 1D mathematical model of GWN, based on MIKE 11, various hydraulic calculations, including proposals of new hydraulic solutions.
Extreme mixing events in rivers
Series: (E-10), 2008, pp.101-107
Although much effort has been expended on quantifying dispersion in simple channels, much less work has focused on compound channels, particularly for over-bank flows. Following an earlier study of a hypothetical compound channel, this paper provides further evidence for the significantly different flow dependence of dispersion coefficients for in-bank and over-bank conditions. The study is based on a natural cross-section of the River Severn, UK, and uses the same method as the earlier work, by combining the Shiono-Knight hydraulic model with the Fischer flow structure integral. The results show that dispersion coefficients are typically two orders of magnitude larger during over-bank flows compared to in-bank flows. Importantly, the maximum dispersion does not occur under the maximum flow. Instead, the maximum dispersion occurs when the flood plain inundation reaches the far edge of the flood plain.
Simulation approach used for the second L-moment derivation of the inverse Gaussian distribution
Series: (E-10), 2008, pp.109-115
Maximum Likelihood (ML) estimate of upper quantiles looses its optimal properties if a wrong distribution is assumed in the ML procedure. Since its estimates base on the main probability mass, the alternative estimation techniques yielding estimates more dependent on upper tail elements of a sample are of interest in flood frequency analysis (FFA). Several systems of describing the shape of probability distribution have been developed and used for matching the assumed distribution to the data. One of them is the system basing on the linear moments. The L-moment estimates have highly desirable properties, like small bias and no algebraic bound of L-moment estimate ratios. It is shown how to use the L-moment system for probability distribution description if the analytical formulas of the linear moments have not been derived. The inverse Gaussian distribution serves as an example.
Sediment transport in the Middle Odra river. Verification of Ackers-White’s formulae with reference to big flows
Series: (E-10), 2008, pp.117-126
The paper presents a method of sediment calculation for a compound river cross-section. An essential point for each calculation of sediment is to keep it in relation to the water flow condition. In the paper, a method is shown in which the verification of sediment calculation can be based upon well-defined Chezy constant.
Uncertainty in the relationship between flow and parameters in models of pollutant transport
Series: (E-10), 2008, pp.127-138
A fluorescent dye-tracer study is usually performed under steady-state flow conditions. However, the model parameters, estimated using the tracer data, depend on the discharges. This paper investigates the uncertainty of the relationship between discharges and parameters of a transient storage and an aggregated dead zone model. We apply a Bayesian statistical approach to derive the cumulative distribution of a range of model parameters conditioned on discharges. The set of tracer data consists of eighteen tracer concentration profiles taken at two cross-sections from the Murray Burn, a stream flowing through the Heriot-Watt University Campus at Riccarton in Edinburgh.
Developments in floodplain inundation modelling
Series: (E-10), 2008, pp.139-149
The advantage of using 2D hydrodynamic models to predict floodplain inundation is that they provide estimates of the velocity field in addition to estimates of flood inundation extent and depth. However, velocity predictions require topographic features on the floodplain to be accurately represented within the solution grid of the numerical model. For a typical developed floodplain this requires a grid resolution of between 2 to 5 m. Floodplain models using this level of resolution are relatively demanding of computer resources and it can take days to simulate a typical 12 hour flood event using a standard dual processor PC. An important issue facing the modeller is therefore, can the floodplain velocity field be predicted to an acceptable level of accuracy using less computer resources. The following presents an initial assessment of two possible methods for achieving this.
Firstly, results from Coarse Grid Models (CGM) at grid resolutions of 10 m and 50 m are compared against benchmark results from a Fine Grid Model (FGM) with a 2 m grid resolution using statistics based on the timing of inundation throughout the modelled domain. Methods for improving the coarse grid predictions using simulations from a limited number of fine grid simulations are presented and illustrated by application to a case study site.
Secondly, the study investigates replicating water level and velocity predictions using a non-linear v-Support Vector Regression (v-SVR) model. This is an integration of v-SVR with CGM predictions using a small number of FGM runs to train v-SVR. The simulated results suggest that the proposed method is able to achieve good predictive results (water level and velocity) as well as provide considerable savings in computer time.
Stochastic Transfer Function simulator of a 1-D flow routing
Series: (E-10), 2008, pp.151-160
The paper presents an application of a Stochastic Transfer Function (STF) approach and a State Dependent Parameter (SDP) transformation of model variables to the combined reservoir management and flow routing problem on the Upper Narew River, NE Poland. The management objective is to reach required flow conditions in the reaches of an ecologically valuable river. A 1-D distributed flow routing model was designed for the study. However, both optimisation methods and reservoir management analysis require numerous model realizations which are computationally very expensive. A much more efficient solution consists of the application of a simplified STF simulator of river flow, which is calibrated on historical data and distributed model realizations for the parts of the river where the observations are not available. The model is stochastic, enabling derivation of prediction uncertainty in a straightforward manner. The obtained optimal control policy is tested on a fully distributed model.
On the rationale of seasonal approach to flood frequency analysis
Series: (E-10), 2008, pp.161-173
The annual peak-flow series of Polish rivers are mixtures of summer and winter flows. The seasonal series are believed to be more homogeneous than the annual series in respect of the distribution. Consequently, a seasonal maxima approach to the stationary flood frequency analysis (FFA) has been introduced by the Polish Hydrological Service. The FFA is performed for the winter and summer peaks separately and the time-dependent annual peak flow quantiles are computed by an union formula of two independent events. Assuming that the seasonal maxima follow the Gumbel distribution, the sampling properties of an annual maxima (AM) quantile estimate got by using the seasonal maxima are examined and compared with those got from the annual maxima (AM) samples. The asymptotic variance and bias of ML-estimate of the upper quantiles are compared with sampling experiment assessment.
Evaluation of the Nida river main current below the perpendicular flood channel outlet
Series: (E-10), 2008, pp.175-185
The development of computational techniques increases the number of easy-to-use software designed for creating mathematical models, including the free CCHE2D software. This encouraged the authors to produce a computer model of a river section influenced by a projected channel outlet conducting water from a small reservoir. A few discharge configurations were tested to verify the potential of the inflowing stream and the jets created in the Nida River. Velocity distribution of water flowing into the Nida River during annual average flow conditions was measured.
Influence of vegetated floodplains on compound channels discharge capacity in 1D modelling
Series: (E-10), 2008, pp.187-194
Three different types of models were applied to compare an impact of floodplain treatment in 1D modelling on compound channels discharge capacity and retention volume of vegetated floodplains. The tested models are based on 1D St Venant equations with the Darcy-Weisbach friction law. A traditional way in which floodplains in 1D modelling are considered storage areas was compared to a model with conveyance of vegetated floodplain and a model with lateral shear stress between the channel and floodplain section, proposed in the Pasche approach. The models were applied to a steady flow in a 50 km long double trapezoidal channel, and differences in rating curves, retention volume of vegetated floodplains, and discharge distribution in a cross-section, were found between the models.
Building wall boundary condition in mathematical modelling of built-up area rapid inundation
Series: (E-10), 2008, pp.195-202
The paper concerns mathematical modelling of rapid flooding flow in urban (built-up) area. The two dimensional Saint Venant equations are assumed as the model of free surface water flow. The model equations are solved using finite volume method. The mass and momentum fluxes are computed applying the Roe scheme of Godunov problem solution. The built-up area is exactly represented in city inundations simulation. Each separate building is excluded from the computational domain. The numerical mesh is generated in flow area between buildings only. An influence of the type of boundary conditions imposed on buildings walls on simulation results is investigated in the paper. The results of numerical computations are examined against laboratory measurements. The laboratory experiment carried out in hydraulic laboratory of the Gdańsk University of Technology is presented in the article.