Changeset 5355
 Timestamp:
 May 22, 2008, 10:52:49 AM (16 years ago)
 Location:
 anuga_work/publications/anuga_2007
 Files:

 2 edited
Legend:
 Unmodified
 Added
 Removed

anuga_work/publications/anuga_2007/anugabibliography.bib
r5354 r5355 1 @INPROCEEDINGS{ modsim2005,1 @INPROCEEDINGS{Nielsen2005, 2 2 AUTHOR = {O. Nielsen and S. Roberts and D. Gray and A. McPherson and A. Hitchman}, 3 3 TITLE = {Hydrodynamic modelling of coastal inundation}, 
anuga_work/publications/anuga_2007/anuga_validation.tex
r5353 r5355 16 16 17 17 % Local LaTeX commands 18 \newcommand{\ANUGA}{ANUGA} %{\textsc{ANUGA}}19 18 %\newcommand{\Python}{\textsc{Python}} 20 19 %\newcommand{\VPython}{\textsc{VPython}} … … 76 75 communities. Geoscience Australia and the Australian National 77 76 University have developed a hydrodynamic inundation modelling tool 78 called \ANUGA{}to help simulate the impact of these hazards.79 The core of \ANUGA{}is a Python implementation of a finitevolume method77 called ANUGA to help simulate the impact of these hazards. 78 The core of ANUGA is a Python implementation of a finitevolume method 80 79 for solving the conservative form of the Shallow Water Wave equation. 81 80 82 In this paper, a number of tests are performed to validate \ANUGA{}. These tests81 In this paper, a number of tests are performed to validate ANUGA. These tests 83 82 range from benchmark problems to wave and flume tank examples. 84 \ANUGA{}is available as Open Source to enable83 ANUGA is available as Open Source to enable 85 84 free access to the software and allow the scientific community to 86 85 use, validate and contribute to the software in the future. … … 128 127 Australia in collaboration with the Mathematical Sciences Institute, 129 128 Australian National University, is developing a software application 130 called \ANUGA{}to model the hydrodynamics of floods, storm surges and129 called ANUGA to model the hydrodynamics of floods, storm surges and 131 130 tsunami. These hazards are modelled using the conservative shallow 132 131 water equations which are described in section~\ref{sec:model}. In 133 \ANUGA{}these equations are solved using a finite volume method as132 ANUGA these equations are solved using a finite volume method as 134 133 described in section~\ref{sec:model}. A more complete discussion of the 135 method can be found in \citet{ modsim2005} where the model and solution134 method can be found in \citet{Nielsen2005} where the model and solution 136 135 technique is validated on a standard tsunami benchmark data set 137 or in \citet{Roberts2007} where parallelisation of ANUGA is discussed. 136 or in \citet{Roberts2007} where the numerical method and parallelisation 137 of ANUGA is discussed. 138 138 This modelling capability is part of 139 139 Geoscience Australia's ongoing research effort to model and 140 140 understand the potential impact from natural hazards in order to 141 141 reduce their impact on Australian communities \citep{Nielsen2006}. 142 \ANUGA{}is currently being trialled for flood142 ANUGA is currently being trialled for flood 143 143 modelling \citep{Rigby2008}. 144 144 145 145 The validity of other hydrodynamic models have been reported 146 elsewhere, with Hubbard and Dodd \citep{Hubbard02} providing an146 elsewhere, with \citet{Hubbard02} providing an 147 147 excellent review of 1D and 2D models and associated validation 148 148 tests. They described the evolution of these models from fixed, nested … … 150 150 moving shoreline. They highlighted the difficulty in verify the 151 151 nonlinear shallow water equations themselves as the only standard 152 analytical solution is that of Carrier and Greenspan153 \citep{Carrier58} that is strictly fornonbreaking waves. Further,154 whilst there is a 2D analytic solution from Thacker (1981), it appears152 analytical solution is that of \citet{Carrier58} that is strictly for 153 nonbreaking waves. Further, 154 whilst there is a 2D analytic solution from \citet{Thacker81}, it appears 155 155 that the circular island wave tank example of Briggs et al will become 156 156 the standard data set to verify the equations. 157 157 158 158 This paper will describe the validation outputs in a similar way to 159 Hubbard and Dodd \citep{Hubbard02} to159 \citet{Hubbard02} to 160 160 present an exhaustive validation of the numerical model. 161 161 Further to these tests, we will … … 164 164 \begin{itemize} 165 165 \item Verification against the 1D analytical solution of Carrier and 166 Greenspan ( Section \ref{sec:carrier})167 \item Testing against 1D (flume) data sets to verify wave height and velocity168 (Section \ref{sec:stage and velocity})166 Greenspan (p~\pageref{sec:carrier}) 167 \item Testing against 1D (flume) data sets to verify wave height and 168 velocity (p~\pageref{sec:stage and velocity}) 169 169 \item Determining friction values from 1D flume data sets 170 (Section \ref{sec:friction}) 171 \item Validation against a genuinely 2D analytical solution of the 172 model equations 173 (Section \ref{sec:XXX}) 170 (p~\pageref{sec:friction}) 171 \item Validation against a genuinely 2D analytical 172 solution of the model equations (p~\ref{sec:XXX}) 174 173 \item Testing against the 2D Okushiri benchmark problem 175 ( Section \ref{sec:okushiri})174 (p~\pageref{sec:okushiri}) 176 175 \item Testing against the 2D data sets modelling wave runup around a circular island by Briggs et al. 177 ( Section \ref{sec:circular island})176 (p~\pageref{sec:circular island}) 178 177 \end{itemize} 179 178 … … 181 180 Throughout the paper, qualitative comparisons will be drawn against 182 181 other models. Moreover, all source code necessary to reproduce the 183 results reported in this paper is available as part of the \ANUGA{}182 results reported in this paper is available as part of the ANUGA 184 183 distribution in the form of a test suite. It is thus possible for 185 184 anyone to readily verify that the implementation meets the … … 187 186 188 187 189 %Hubbard and Dodd's model, OTT2D, has some similarities to \ANUGA{}, and188 %Hubbard and Dodd's model, OTT2D, has some similarities to ANUGA, and 190 189 %whilst the mesh can be refined, it is based on rectangular mesh. 191 190 192 The \ANUGA{} model and numerical scheme is briefly described in 193 section~\ref{sec:model}. A detailed description of the numerical 194 scheme and software implementation can be found in the MODSIM, CTAC 195 etc papers. The six case studies to validation and verify \ANUGA{} 191 %The ANUGA model and numerical scheme is briefly described in 192 %section~\ref{sec:model}. A more detailed description of the numerical 193 %scheme and software implementation can be found in \citet{Nielsen2005} and 194 %\citet{Roberts2007}. 195 The six case studies to validation and verify ANUGA 196 196 will be presented in section~\ref{sec:validation}, with the 197 197 conclusions outlined in section~\ref{sec:conclusions}. … … 208 208 \citet{Stoker57} and \citet{Peregrine67} for the background or 209 209 \citet{Roberts1999} for more details on the mathematical model 210 used by \ANUGA{}.210 used by ANUGA. 211 211 212 212 The conservation form of the shallow water wave 213 equations used in \ANUGA{}are:213 equations used in ANUGA are: 214 214 \[ 215 215 \frac{\partial \UU}{\partial t}+\frac{\partial \EE}{\partial … … 258 258 %%know it is excellent? 259 259 260 \ANUGA{}uses a finitevolume method as260 ANUGA uses a finitevolume method as 261 261 described in \citet{Roberts2007} where the study area is represented by an 262 262 unstructured triangular mesh in which the vector of conserved quantities … … 268 268 269 269 270 The approach used in \ANUGA{}are distinguished from many270 The approach used in ANUGA are distinguished from many 271 271 other implementations (e.g. \citet{Hubbard02} or \citet{Zhang07}) by the 272 272 following features: … … 278 278 critical flow transitions using one general approach. We have 279 279 found this scheme to be pleasingly simple, robust and efficient. 280 \item \ANUGA{}does not employ a shoreline detection algorithm as the280 \item ANUGA does not employ a shoreline detection algorithm as the 281 281 centralupwind scheme is capable of resolving fluxes arising between 282 wet and dry cells. \ANUGA{}does optionally bypass unnecessary282 wet and dry cells. ANUGA does optionally bypass unnecessary 283 283 computations for drydry cell boundaries for purely performance reasons. 284 \item \ANUGA{}employs a second order spatial reconstruction of triangles284 \item ANUGA employs a second order spatial reconstruction of triangles 285 285 to produce a piecewise linear function construction of the conserved 286 286 quantities. This function is allowed to be discontinuous across the … … 316 316 317 317 318 \ANUGA{}utilises a general velocity limiter described in the318 ANUGA utilises a general velocity limiter described in the 319 319 manual which guarantees a gradual compression of computed velocities 320 320 in the presence of very shallow depths: … … 326 326 327 327 328 \ANUGA{}is mostly written in the objectoriented programming328 ANUGA is mostly written in the objectoriented programming 329 329 language Python with computationally intensive parts implemented 330 330 as highly optimised shared objects written in C. … … 335 335 language syntax. In addition, Python's automatic memory management, 336 336 dynamic typing, object model and vast number of libraries means that 337 \ANUGA scripts can be produced quickly and can be adapted fairly easily to337 ANUGA scripts can be produced quickly and can be adapted fairly easily to 338 338 changing requirements. 339 339 … … 342 342 \section{Validation} 343 343 \label{sec:validation} Validation is an ongoing process and the purpose of this paper 344 is to describe a range of tests that validate \ANUGA{}as a hydrodynamic model.344 is to describe a range of tests that validate ANUGA as a hydrodynamic model. 345 345 This section will describe the six tests outlined in section~\ref{sec:intro}. 346 346 Run times where specified measure the model time only and exclude model setup, … … 358 358 This section will describe tilting flume tank experiments that were 359 359 conducted at the Gordon McKay Hydraulics Laboratory at the University of 360 Queensland that confirm \ANUGA{}'s ability to estimate wave height360 Queensland that confirm ANUGA's ability to estimate wave height 361 361 and velocity. The same flume tank simulations were also used 362 362 to explore Manning's friction and this will be described in the next section. … … 389 389 \begin{figure}[htbp] 390 390 \centerline{\includegraphics[width=4in]{uqflumedepth}} 391 \caption{Comparison of wave tank and \ANUGA{}water height at .4 m391 \caption{Comparison of wave tank and ANUGA water height at .4 m 392 392 from the gate}\label{fig:uqflumedepth} 393 393 \end{figure} … … 395 395 \begin{figure}[htbp] 396 396 \centerline{\includegraphics[width=4in]{uqflumevelocity}} 397 \caption{Comparison of wave tank and \ANUGA{}water velocity at .45 m397 \caption{Comparison of wave tank and ANUGA water velocity at .45 m 398 398 from the gate}\label{fig:uqflumevelocity} 399 399 \end{figure} … … 450 450 \begin{figure}[htbp] 451 451 \centerline{\includegraphics[width=4in]{uqfrictiondepth}} 452 \caption{Comparison of wave tank and \ANUGA{}water height at .4 m452 \caption{Comparison of wave tank and ANUGA water height at .4 m 453 453 from the gate, simulated using a Mannings friction of 0.0 and 454 454 0.1.}\label{fig:uqfrictiondepth} … … 478 478 \centerline{\includegraphics[width=4in]{ch7.png}} 479 479 \centerline{\includegraphics[width=4in]{ch9.png}} 480 \caption{Comparison of wave tank and \ANUGA{}water stages at gauge480 \caption{Comparison of wave tank and ANUGA water stages at gauge 481 481 5,7 and 9.}\label{fig:val} 482 482 \end{figure} … … 486 486 \centerline{\includegraphics[width=4in]{okushirimodel.jpg}} 487 487 \caption{Complex reflection patterns and runup into Monai Valley 488 simulated by \ANUGA{}and visualised using our netcdf OSG488 simulated by ANUGA and visualised using our netcdf OSG 489 489 viewer.}\label{fig:run} 490 490 \end{figure} 491 491 492 492 The wave tank simulation of the Hokkaido tsunami was used as the 493 first scenario for validating \ANUGA{}. The dataset provided493 first scenario for validating ANUGA. The dataset provided 494 494 bathymetry and topography along with initial water depth and the 495 495 wave specifications. The dataset also contained water depth time 496 496 series from three wave gauges situated offshore from the simulated 497 inundation area. The \ANUGA{}model comprised $41404$ triangles497 inundation area. The ANUGA model comprised $41404$ triangles 498 498 and took about $1330$ s to run on the test platform described in 499 499 Section~\ref{sec:validation}. … … 504 504 505 505 Figure~\ref{fig:val} compares the observed wave tank and modelled 506 \ANUGA{}water depth (stage height) at one of the gauges. The plots507 show good agreement between the two time series, with \ANUGA{506 ANUGA water depth (stage height) at one of the gauges. The plots 507 show good agreement between the two time series, with ANUGA 508 508 closely modelling the initial draw down, the wave shoulder and the 509 509 subsequent reflections. The discrepancy between modelled and … … 511 511 condition in the physical tank not being uniformly zero. Similarly 512 512 good comparisons are evident with data from the other two gauges. 513 Additionally, \ANUGA{}replicates exceptionally well the 32~m Monai513 Additionally, ANUGA replicates exceptionally well the 32~m Monai 514 514 Valley runup, and demonstrates its occurrence to be due to the 515 515 interaction of the tsunami wave with two juxtaposed valleys above … … 517 517 518 518 This successful replication of the tsunami wave tank simulation on a 519 complex 3D beach is a positive first step in validating the \ANUGA{}519 complex 3D beach is a positive first step in validating the ANUGA 520 520 modelling capability. 521 521 … … 570 570 \section{Conclusions} 571 571 \label{sec:conclusions} 572 \ANUGA{}is a flexible and robust modelling system572 ANUGA is a flexible and robust modelling system 573 573 that simulates hydrodynamics by solving the shallow water wave 574 574 equation in a triangular mesh. It can model the process of wetting … … 576 576 capturing hydraulic shocks due to the ability of the finitevolume 577 577 method to accommodate discontinuities in the solution. 578 \ANUGA{}can take as input bathymetric and topographic datasets and578 ANUGA can take as input bathymetric and topographic datasets and 579 579 simulate the behaviour of riverine flooding, storm surge, 580 580 tsunami or even dam breaks. 581 Initial validation using wave tank data supports \ANUGA{}'s581 Initial validation using wave tank data supports ANUGA's 582 582 ability to model complex scenarios. Further validation will be 583 583 pursued as additional datasets become available. 584 The \ANUGA{} source code isavailable584 The ANUGA source code and validation case studies reported here are available 585 585 at \url{http://sourceforge.net/projects/anuga}. 586 586
Note: See TracChangeset
for help on using the changeset viewer.