Changeset 7377


Ignore:
Timestamp:
Aug 16, 2009, 8:28:05 PM (15 years ago)
Author:
ole
Message:

Jane's (and some of mine) comments and suggestions.
See log file and FIXME's

File:
1 edited

Legend:

Unmodified
Added
Removed
  • anuga_work/publications/boxing_day_validation_2008/patong_validation.tex

    r7303 r7377  
    1414%----------title-------------%
    1515\title{Benchmarking Tsunami Models using the December 2004 Indian
    16   Ocean Tsunami and its Impact at Patong Beach}
     16  Ocean Tsunami and its Impact at Patong Bay}
    1717
    1818%-------authors-----------
     
    3535%------Abstract--------------
    3636\begin{abstract}
    37 In this paper a new benchmark for tsunami model validation is
    38 proposed. The benchmark is based upon the 2004 Indian Ocean tsunami,
    39 which affords a uniquely large amount of observational data for model
    40 comparison. Unlike the small number of existing benchmarks, the
     37This paper proposes a new benchmark for tsunami model validation.
     38The benchmark is based upon the 2004 Indian Ocean tsunami,
     39which affords a uniquely large amount of observational data for events of this kind.
     40Unlike the small number of existing benchmarks, the
    4141proposed test validates all three stages of tsunami evolution -
    42 generation, propagation and inundation. Specifically we use geodetic
     42generation (FIXME (Jane): really?), propagation and inundation. Specifically we use geodetic
    4343measurements of the Sumatra--Andaman earthquake to validate the
    4444tsunami source, altimetry data from the \textsc{jason} satellite to
    4545test open ocean propagation, eye-witness accounts to assess near shore
    46 propagation and a detailed inundation survey of Patong Bay, Thailand
     46propagation and a detailed inundation survey of Patong city, Thailand
    4747to compare model and observed inundation. Furthermore we utilise this
    4848benchmark to further validate the hydrodynamic modelling tool
     
    6262
    6363\section{Introduction}
    64 Tsunami are a potential hazard to coastal communities all over the
     64Tsunami is a potential hazard to coastal communities all over the
    6565world. A number of recent large events have increased community and
    6666scientific awareness of the need for effective detection, forecasting,
    67 and emergency preparedness. Probabilistic, geological, hydrodynamic,
    68 and economic models are required to predict the location and
     67and emergency preparedness. Probabilistic, geophysical and hydrodynamic
     68models are required to predict the location and
    6969likelihood of an event, the initial sea floor deformation and
    70 subsequent propagation and inundation of the tsunami, the
    71 effectiveness of hazard mitigation procedures and the economic impact
    72 of such measures and of the event itself. Here we focus on modelling of
    73 the physical processes.
    74 %OLE: I commented this out 23 June 2009 as there was no reference.
    75 %For discussion on economic and decision based
    76 %models refer to~\cite{} and the references therein.
     70subsequent propagation and inundation of the tsunami. Engineering, economic and social vulnerability models can then be used to estimate the
     71impact of the event as well as the effectiveness of hazard mitigation
     72procedures. In this paper, we focus on modelling of
     73the physical processes only.
    7774
    7875Various approaches are currently used to assess the potential impact
     
    126123data also significantly increases the uncertainty of the validation
    127124experiment that may constrain the ability to make unequivocal
    128 statements~\cite{bates01}.
     125statements~\cite{bates01}. FIXME (Jane): Because?
     126FIXME (Phil): references to all of the paragraph above, please
    129127
    130128Currently, the extent of tsunami-related field data is limited. The
     
    133131tsunamis pose greatest threat. The resulting lack of data has limited
    134132the number of field data sets available to validate tsunami
    135 models. Synolakis et al~\cite{synolakis07} have developed a set of
     133models.
     134
     135Synolakis et al~\cite{synolakis07} have developed a set of
    136136standards, criteria and procedures for evaluating numerical models of
    137137tsunami. They propose three analytical solutions to help identify the
     
    151151provides a good test for real-time forecasting models since the tsunami
    152152was recorded at three tsunameters. The test requires matching the
    153 propagation model data with the DART recording to constrain the
     153tsunami propagation model output with the DART recording to constrain the
    154154tsunami source model, and then using it to reproduce the tide gauge
    155 record at Hilo.
     155record at Hilo, Hawaii.
     156FIXME (Jane): Are the tsunameters and the DART recordings the same thing?
    156157
    157158In this paper we develop a field data benchmark to be used in
     
    164165of the tsunami source, altimetry data from the JASON satellite to test
    165166open ocean propagation, eye-witness accounts to assess near shore
    166 propagation, and a detailed inundation survey of Patong Bay, Thailand
     167propagation, and a detailed inundation survey of Patong city, Thailand
    167168to compare model and observed inundation. A description of the data
    168169required to construct the benchmark is given in
     
    170171
    171172An associated aim of this paper is to illustrate the use of this new
    172 benchmark to validate an operational tsunami inundation model called
     173benchmark to validate a dedicated inundation model called
    173174\textsc{anuga} used by Geoscience Australia. A description of
    174175\textsc{anuga} is given in Section~\ref{sec:models} and the validation
    175176results are given in Section~\ref{sec:results}.
    176177
    177 The numerical models used to model tsunami are extremely
    178 computationally intensive. Full resolution models of the entire
     178The numerical models used to simulate tsunami impact
     179are computationally intensive and high resolution models of the entire
    179180evolution process will often take a number of days to
    180 run. Consequently the uncertainty in model predictions is difficult to
    181 quantify. However model uncertainty should not be ignored. Section
    182 ~\ref{sec:sensitivity} provides a simple sensitivity analysis that can
     181run. Consequently, the uncertainty in model predictions is difficult to
     182quantify as it would require a very large number of runs.
     183However, model uncertainty should not be ignored. Section
     184~\ref{sec:sensitivity} provides a simple analysis that can
    183185be used to investigate the sensitivity of model predictions to model
    184186parameters.
     
    192194seismometers, tide gauges, \textsc{gps} surveys, satellite overpasses,
    193195subsequent coastal field surveys of run-up and flooding, and
    194 measurements of coseismic displacements and bathymetry from ship-based
     196measurements of coseismic displacements as well as bathymetry from ship-based
    195197expeditions, have now been made
    196 available. %~\cite{vigny05,amnon05,kawata05,liu05}.
     198available. %~\cite{vigny05,amnon05,kawata05,liu05}. FIXME (Ole): Refs?
    197199In this section we present the corresponding data necessary to implement
    198200the proposed benchmark for each of the three stages of the tsunami's evolution.
     
    208210by the 2004 Sumatra--Andaman earthquake.
    209211
    210 The 2004 Sumatra--Andaman tsunami was generated by a severe coseismic
    211 displacement of the sea floor as a result of one of the largest
     212The 2004 Sumatra--Andaman tsunami was generated by a coseismic
     213displacement of the sea floor resulting from one of the largest
    212214earthquakes on record. The mega-thrust earthquake started on the 26
    213215December 2004 at 0h58'53'' UTC (or just before 8 am local time)
     
    248250
    249251\subsection{Propagation}
     252\label{sec:propagation data}
    250253Once generated, a tsunami will propagate outwards from the source until
    251254it encounters the shallow water bordering coastal regions. This period
     
    258261
    259262\subsubsection{Bathymetry Data}
    260 A number of raw data sets were obtained, analysed and checked for
    261 quality and subsequently gridded for easier visualisation and input
    262 into the tsunami models. The resulting grid data is relatively coarse
    263 in the deeper water and becomes progressively finer as the distance to
    264 Patong Bay decreases.
    265 
    266 The nested bathymetry grid was generated from:
     263The bathymetry data used in this study was derived from the following
     264sources:
    267265\begin{itemize}
    268266\item a two arc minute grid data set covering the Bay of Bengal,
    269267  DBDB2, obtained from US Naval Research Labs;
    270268\item a 3 second arc grid covering the whole of the Andaman Sea based
    271   on Thai Navy charts no. 45 and no. 362; and
     269  on Thai Navy charts no. 45 and no. 362; and 
    272270\item a one second grid created from the digitised Thai Navy
    273271  bathymetry chart, no. 358, which covers Patong Bay and the
    274   immediately adjacent regions.
     272  immediately adjacent regions. (FIXME (Ole): How was the grid created from these digitised points?)
    275273\end{itemize}
    276 
    277 The final bathymetry data set consists of four nested grids obtained
    278 via interpolation and resampling of the aforementioned data sets. The
    279 four grids are shown in Figure~\ref{fig:nested_grids}.  The coarsest
     274FIXME (Jane): Refs for all these.
     275
     276%A number of raw data sets were obtained, analysed and checked for
     277%quality and subsequently gridded for easier visualisation and input
     278%into the tsunami models.
     279
     280These sets were combined via
     281interpolation and resampling to produce four nested grids
     282which are relatively coarse in the deeper water and
     283progressively finer as the distance to
     284Patong Beach decreases as shown in Figure~\ref{fig:nested_grids}. 
     285
     286The coarsest
    280287bathymetry was obtained by interpolating the DBDB2 grid to a 27 second
    281288arc grid. A subsection of this region was then replaced by nine second
    282289data which was generated by sub-sampling the three second of arc grid from
    283 NOAA. A subset of the nine second grid was replaced by the three second
     290NOAA (FIXME (Jane): This was not mentioned in the dots above).
     291A subset of the nine second grid was replaced by the three second
    284292data. Finally, the one second grid was used to approximate the
    285293bathymetry in Patong Bay and the immediately adjacent regions. Any
    286294points that deviated from the general trend near the boundary were
    287 deleted.
     295deleted as a quality check.
    288296
    289297The sub-sampling of larger grids was performed by using {\bf resample},
     
    300308\begin{center}
    301309\includegraphics[width=0.75\textwidth,keepaspectratio=true]{nested_grids}
    302 \caption{Nested grids of the elevation data.}
     310\caption{Nested bathymetry grids.}
    303311\label{fig:nested_grids}
    304312\end{center}
     
    306314
    307315\subsubsection{JASON Satellite Altimetry}\label{sec:data_jason}
    308 During the 26 December 2004 event, the Jason satellite tracked from
     316During the 26 December 2004 event, the \textsc{jason} satellite tracked from
    309317north to south and over the equator at 02:55 UTC nearly two hours
    310318after the earthquake \cite{gower05}. The satellite recorded the sea
    311319level anomaly compared to the average sea level from its previous five
    312 passes over the same region in the 20-30 days prior.  This data was
     320passes over the same region in the 20-30 days prior. This data was
    313321used to validate the propagation stage in Section
    314322\ref{sec:resultsPropagation}.
     
    334342
    335343\subsection{Inundation}
     344\label{sec:inundation data}
     345FIXME (Ole): Technically propagation covers everything up to
     346the coastline and inundation everything on-shore.
     347This means that ANUGA covers the final part of the propagation and the inundation part. Should we adopt this distiction throughout the paper?
     348
    336349Inundation refers to the final stages of the evolution of a tsunami and
    337 covers the propagation of the tsunami in shallow coastal water and the
     350covers the propagation of the tsunami in coastal waters and the
    338351subsequent run-up onto land. This process is typically the most
    339352difficult of the three stages to model due to thin layers of water
     
    343356data which is often not available. In the case of model validation
    344357high quality field measurements are also required. For the proposed
    345 benchmark the authors have obtained a high resolution bathymetry and
     358benchmark a high resolution bathymetry (FIXME (Ole): Bathymetry ?) and
    346359topography data set and a high quality inundation survey map from the
    347 CCOP in Thailand (\cite{szczucinski06}) to validate model inundation.
    348 
    349 The datasets necessary for reproducing the results
    350 of the inundation stage are available on Sourceforge under the \textsc{anuga}
    351 project (\url{http://sourceforge.net/projects/anuga}). At the time of
    352 writing the direct link is \url{http://tinyurl.com/patong2004-data}.
    353 %
    354 %\url{http://sourceforge.net/project/showfiles.php?group_id=172848&package_id=319323&release_id=677531}.
    355 
    356 In this section we also present eye-witness accounts which can be used to qualitatively validate tsunami inundation.
     360Coordinating Committee Co-ordinating Committee for Geoscience Programmes
     361in East and Southeast Asia (CCOP) (\cite{szczucinski06}) was obtained
     362to validate model inundation. See also acknowledgements at the end of this paper.
     363
     364In this section we also present eye-witness accounts which can be used
     365to qualitatively validate tsunami inundation.
    357366
    358367\subsubsection{Topography Data}
    359368A one second grid was used to approximate the topography in Patong
    360369Bay. This elevation data was again created from the digitised Thai
    361 Navy bathymetry chart, no 358. A visualisation of the elevation data
     370Navy bathymetry chart, no 358.
     371FIXME (Ole): I don't think so. The Navy chart is only offshore.
     372
     373 A visualisation of the elevation data
    362374set used in Patong Bay is shown in
    363 Figure~\ref{fig:patong_bathymetry}. The continuous topography is an
    364 interpolation of known elevation measured at the coloured dots.
     375Figure~\ref{fig:patong_bathymetry}. The continuous topography
     376(FIXME(Jane): What is meant by this?) is an
     377interpolation of known elevation measured at the coloured dots. FIXME ??
    365378
    366379\begin{figure}[ht]
    367380\begin{center}
    368381\includegraphics[width=8.0cm,keepaspectratio=true]{patong_bay_data.jpg}
    369 \caption{Visualisation of the elevation data set used in Patong Bay showing data points, contours, rivers and roads draped over the final model.}
     382\caption{3D visualisation of the elevation data set used in Patong Bay showing data points, contours, rivers and roads draped over the final model.}
    370383\label{fig:patong_bathymetry}
    371384\end{center}
    372385\end{figure}
     386FIXME (Jane): legend? Were the contours derived from the final dataset?
     387This is not the entire mode, only the bay and the beach.
    373388
    374389\subsubsection{Buildings and Other Structures}
    375 Human-made build and structures can significantly effect tsunami
    376 inundation. The location and size and number of floors of the
    377 buildings in Patong Bay were extracted from a GIS data set provided by
    378 the CCOP in Thailand (see acknowledgements at the end of this paper).
     390Human-made buildings and structures can significantly affect tsunami
     391inundation. The footprint and number of floors of the
     392buildings in Patong Bay were extracted from a GIS data set which was also provided by the CCOP (see Section \ref{sec:inundation data} for details).
    379393The heights of these
    380 buildings were estimated assuming that each floor has a height of 3 m.
     394buildings were estimated assuming that each floor has a height of 3 m and they
     395were added to the topographic dataset.
    381396
    382397\subsubsection{Inundation Survey}
    383 Tsunami run-up is often the cause of the largest financial and human
     398Tsunami run-up is the cause of the largest financial and human
    384399losses, yet run-up data that can be used to validate model run-up
    385 predictions is scarce. Of the two field benchmarks proposed in~\cite{synolakis07},
     400predictions is scarce. Of the two field benchmarks proposed
     401in~\cite{synolakis07},
    386402only the Okushiri benchmark facilitates comparison between
    387403modelled and observed run-up. One of the major strengths of the
     
    390406rather than at a series of discrete sites. The survey map is
    391407shown in Figure~\ref{fig:patongescapemap} and plots the maximum run-up
    392 of the 2004 tsunami in Patong Bay. Refer to Szczucinski et
     408of the 2004 Indian Ocean tsunami in Patong city. Refer to Szczucinski et
    393409al~\cite{szczucinski06} for further details.
     410
     411\begin{figure}[ht]
     412\begin{center}
     413\includegraphics[width=8.0cm,keepaspectratio=true]{patongescapemap.jpg}
     414\caption{Tsunami survey mapping the maximum observed inundation at
     415  Patong beach courtesy of the CCOP \protect \cite{szczucinski06}.}
     416\label{fig:patongescapemap}
     417\end{center}
     418\end{figure}
     419
    394420
    395421\subsubsection{Eyewitness Accounts}\label{sec:eyewitness data}
     
    402428minutes after the source rupture (09:55am to 10:05am local time).
    403429
    404 Two videos were sourced from the internet\footnote{The footage is
     430Two videos were sourced\footnote{The footage is
    405431widely available and can for example be obtained from
    406432\url{http://www.archive.org/download/patong_bavarian/patong_bavaria.wmv}
     
    410436%http://wizbangblog.com/content/2005/01/01/wizbang-tsunami.php
    411437which include footage of the tsunami in Patong Bay on the day
    412 of the Indian Ocean Tsunami. Both videos show an already inundated
     438of the 2004 Indian Ocean Tsunami. Both videos show an already inundated
    413439group of buildings. They also show what is to be assumed as the second
    414440and third waves approaching and further flooding of the buildings and
     
    438464were found to be in the range of 5 to 7 metres per second (+/- 2 m/s)
    439465in the north and 0.5 to 2 metres per second (+/- 1 m/s) in the south.
     466FIXME (Jane): How were these error bounds derived?
    440467Water depths could also
    441468be estimated from the videos by the level at which water rose up the
     
    446473speeds in the range of 2 to 5 m/s.
    447474
    448 \begin{figure}[ht]
    449 \begin{center}
    450 \includegraphics[width=8.0cm,keepaspectratio=true]{patongescapemap.jpg}
    451 \caption{Tsunami survey mapping the maximum observed inundation at
    452   Patong beach courtesy of the Thai Department of Mineral Resources
    453   \protect \cite{szczucinski06}.}
    454 \label{fig:patongescapemap}
    455 \end{center}
    456 \end{figure}
    457475
    458476\subsection{Validation Check-List}
     
    465483 \item reproduce the vertical deformation observed in north-western
    466484   Sumatra and along the Nicobar--Andaman islands (see
    467    Section~\ref{sec:gen_data}).
     485   Section~\ref{sec:gen_data}),
    468486 \item reproduce the \textsc{jason} satellite altimetry sea surface
    469    anomalies (see Section~\ref{sec:data_jason}).
    470  \item reproduce the inundation survey map in Patong bay
    471    (Figure~\ref{fig:patongescapemap}).
     487   anomalies (see Section~\ref{sec:data_jason}),
     488 \item reproduce the inundation survey map in Patong city
     489   (Figure~\ref{fig:patongescapemap}),
    472490 \item simulate a leading depression followed by two distinct crests
    473    of decreasing magnitude.
     491   of decreasing magnitude at the beach, and
    474492 \item predict the water depths and flow speeds, at the locations of
    475493   the eye-witness videos, that fall within the bounds obtained from
     
    479497Ideally, the model should also be compared to measured timeseries of
    480498waveheights and velocities but the authors are not aware of the
    481 availability of such data.
     499availability of such data near Patong Bay.
    482500
    483501
     
    493511
    494512There are various approaches to modelling the expected crustal
    495 deformation from an earthquake at depth. Most approaches model the
     513deformation from an earthquake. Most approaches model the
    496514earthquake as a dislocation in a linear elastic medium. Here we use
    497515the method of Wang et al~\cite{wang03}. One of the main advantages
     
    509527using a combination of Hankel's transform and Wang et al's
    510528implementation of the Thomson-Haskell propagator
    511 algorithm~\cite{wang03}. Once the Green's functions are calculated we
    512 use a slightly modified version of \textsc{edcmp} to calculate the sea
     529algorithm~\cite{wang03}. Once the Green's functions are calculated
     530a slightly modified version of \textsc{edcmp}\footnote{For this study,
     531we have made minor modifications
     532to \textsc{edcmp} in order for it to provide output in a file format
     533compatible with the propagation code in the following section. Otherwise it
     534is similar to the original code.} is used to calculate the sea
    513535floor deformation for a specific subfault. This second code
    514536discretises the subfault into a set of unit sources and sums the
     
    517539deformation caused by a two dimensional dislocation along the
    518540subfault. This step is possible because of the linearity of the
    519 governing equations. For this study, we have made minor modifications
    520 to \textsc{edcmp} in order for it to provide output in a file format
    521 compatible with the propagation code in the following section. Otherwise it
    522 is similar to the original code.
    523 
    524 In order to calculate the crustal deformation using these codes we
    525 need a model that describes the variation in elastic
     541governing equations.
     542
     543In order to calculate the crustal deformation using these codes
     544a model that describes the variation in elastic
    526545properties with depth and a slip model of the earthquake to describe
    527 the dislocation. The elastic parameters used for this study are the
    528 same as those in Table 2 of Burbidge~\cite{burbidge08}. For the slip
     546the dislocation is required.
     547The elastic parameters used for this study are the
     548same as those in Table 2 of Burbidge et al~\cite{burbidge08}. For the slip
    529549model, there are many possible models for the 2004 Andaman--Sumatran
    530550earthquake to select from
     
    532552determined from various geological surveys of the site. Others solve
    533553an inverse problem which calibrates the source based upon the tsunami
    534 wave signal, the seismic signal and/or the run-up. The source
     554wave signal, the seismic signal and/or even the run-up.
     555The source
    535556parameters used here to simulate the 2004 Indian Ocean tsunami were
    536557taken from the slip model G-M9.15 of Chlieh
     
    550571
    551572\subsection{Propagation}\label{sec:modelPropagation}
    552 We use the \textsc{ursga} model described below to simulate the
    553 propagation of the 2004 tsunami in the deep ocean ocean, based on a
     573The \textsc{ursga} model described below was used to simulate the
     574propagation of the 2004 Indian Ocean tsunami across the open ocean, based on a
    554575discrete representation of the initial deformation of the sea floor, as
    555576described in Section~\ref{sec:modelGeneration}. For the models shown
    556 here, we assume that the uplift is instantaneous and creates a wave of
     577here, the uplift is assumed to be instantaneous and creates a wave of
    557578the same size and amplitude as the co-seismic sea floor deformation.
    558579
     
    563584spherical co-ordinates with friction and Coriolis terms. The code is
    564585based on Satake~\cite{satake95} with significant modifications made by
    565 the \textsc{urs} corporation~\cite{thio08} and Geoscience
    566 Australia~\cite{burbidge08}. The tsunami is propagated via the nested
    567 grid system. Coarse grids are used in the open ocean and the finest
    568 resolution grid is employed in the region of most
    569 interest. \textsc{Ursga} is not publicly available.
     586the \textsc{urs} corporation, Thio et al~\cite{thio08} and Geoscience
     587Australia, Burbidge et al~\cite{burbidge08}.
     588The tsunami was propagated via the nested
     589grid system described in Section \ref{sec:propagation data} where
     590the coarse grids were used in the open ocean and the finest
     591resolution grid was employed in the region closest to Patong bay.
     592\textsc{Ursga} is not publicly available.
    570593
    571594\subsection{Inundation}\label{sec:modelInundation}
     
    578601Geoscience Australia tsunami modelling methodology is based on a
    579602hybrid approach using models like \textsc{ursga} for tsunami
    580 propagation up to a 100 m depth contour.
     603propagation up to an offshore depth contour, typically 100 m.
    581604%Specifically we use the \textsc{ursga} model to simulate the
    582605%propagation of the 2004 Indian Ocean tsunami in the deep ocean, based
    583606%on a discrete representation of the initial deformation of the sea
    584607%floor, described in Section~\ref{sec:modelGeneration}.
    585 The wave signal is then used as a time varying boundary condition for
     608The wave signal and the velocity field is then used as a
     609time varying boundary condition for
    586610the \textsc{anuga} inundation simulation.
    587611% A description of \textsc{anuga} is the following section.
    588612
    589613\subsubsection{ANUGA}
    590 \textsc{Anuga} is an Open Source hydrodynamic inundation tool that
     614\textsc{Anuga} is a Free and Open Source hydrodynamic inundation tool that
    591615solves the conserved form of the depth-integrated nonlinear shallow
    592 water wave equations. The scheme used by \textsc{anuga}, first
     616water wave equations using a Finite-Volume scheme on an
     617unstructured triangular mesh.
     618The scheme, first
    593619presented by Zoppou and Roberts~\cite{zoppou99}, is a high-resolution
    594620Godunov-type method that uses the rotational invariance property of
     
    598624et al~\cite{kurganov01} for solving one-dimensional conservation
    599625equations. The numerical scheme is presented in detail in
    600 Roberts and Zoppou~\cite{zoppou99,roberts00} and
     626Roberts and Zoppou~\cite{zoppou00,roberts00} and
    601627Nielsen et al~\cite{nielsen05}. An important capability of the
    602 software is that it can model the process of wetting and drying as
    603 water enters and leaves an area. This means that it is suitable for
     628finite-volume scheme is that discontinuities in all conserved quantities
     629are allowed at every edge in the mesh. This means that the tool is
     630well suited to adequately resolving hydraulic jumps, transcritical flows and
     631the process of wetting and drying. This means that \textsc{Anuga}
     632is suitable for
    604633simulating water flow onto a beach or dry land and around structures
    605 such as buildings. It is also capable of adequately resolving
    606 hydraulic jumps due to the ability of the finite-volume method to
    607 handle discontinuities. The numerical scheme can also handle
    608 transitions between sub-critical and super-critical flow regimes
    609 seamlessly. \textsc{Anuga} has been validated against a number of
    610 analytical solutions and the wave tank simulation of the 1993 Okushiri
     634such as buildings. \textsc{Anuga} has been validated against
     635%a number of analytical solutions and  FIXME: These have not been published
     636the wave tank simulation of the 1993 Okushiri
    611637Island tsunami~\cite{nielsen05,roberts06}.
     638FIXME (Ole): Add reference to Tom Baldock's Dam Break valiadation of ANUGA.
     639
    612640
    613641%================Section===========================
     
    629657(arrows point down) during and immediately after the earthquake. Most
    630658of this data comes from uplifted or subsided coral heads. The length of
    631 vector increases with the magnitude of the displacement; the length
     659the vector increases with the magnitude of the displacement; the length
    632660corresponding to 1 m of observed motion is shown in the top right
    633661corner of the figure. As can be seen, the source model detailed in
     
    642670points) is only 0.06 m, well below the typical error of the
    643671observations of between 0.25 and 1.0 m. However, the occasional point
    644 has quite a large error (over 1 m); for example a couple
    645 uplifted/subsided points appear to be on a wrong side of the predicted
     672has quite a large error (over 1 m); for example a couple of
     673uplifted/subsided points appear to be on a wrong
     674(FIXME (Jane): This is incorrect) side of the predicted
    646675pivot line~\ref{fig:surface_deformation}. The excellence of the fit is
    647676not surprising, since the original slip model was chosen
    648677by~\cite{chlieh07} to fit this (and the seismic data) well.
    649678This does demonstrate, however, that \textsc{edgrn} and our modified version of
    650 \textsc{edstat} can reproduce the correct pattern of vertical
     679\textsc{edstat} (FIXME(Jane): This has never been mentioned before)
     680can reproduce the correct pattern of vertical
    651681deformation very well when the slip distribution is well constrained
    652682and when reasonable values for the elastic properties are used.
     
    665695  hand corner of the figure. The cross marks show the location of
    666696  the pivot line (the region between the uplift and subsided region
    667   where the uplift is zero) derived from remote sensing. All the
     697  where the uplift is zero) derived from remote sensing
     698  (FIXME(Jane): How was that possible?). All the
    668699  observational data are from the dataset collated
    669700  by~\cite{chlieh07}.}
     
    686717shown in Figure~\ref{fig:computational_domain}.
    687718
     719\begin{figure}[ht]
     720\begin{center}
     721%\includegraphics[width=5.0cm,keepaspectratio=true]{extent_of_ursga_model.jpg}
     722%\includegraphics[width=5.0cm,keepaspectratio=true]{ursgaDomain.jpg}
     723\includegraphics[width=5.0cm,keepaspectratio=true]{extent_of_ANUGA_model.jpg}
     724\caption{Computational domain of the \textsc{ursga} simulation (inset: white and black squares and main: black square) and the \textsc{anuga} simulation (main and inset: red polygon).}
     725\label{fig:computational_domain}
     726\end{center}
     727\end{figure}
     728
     729
    688730Figure \ref{fig:jasonComparison} provides a comparison of the
    689 \textsc{ursga}-predicted sea surface elevation with the JASON
     731\textsc{ursga}-predicted sea surface elevation with the \textsc{jason}
    690732satellite altimetry data. The \textsc{ursga} model replicates the
    691733amplitude and timing of the the wave observed at $2.5^0$ South,
     
    696738as can be seen in the satellite data. Also note
    697739that the \textsc{ursga} model prediction of the ocean surface
    698 elevation becomes out of phase with the JASON data at $3^0$ to $7^0$ North
     740elevation becomes out of phase with the \textsc{jason}
     741data at $3^0$ to $7^0$ North
    699742latitude. Chlieh et al~\cite{chlieh07} also observed these misfits and
    700743suggest it is caused by a reflected wave from the Aceh Peninsula that
    701744is not resolved in the model due to insufficient resolution of the
    702745computational mesh and bathymetry data. This is also a limitation of
    703 the model presented here, but probably could be improved by nesting
     746the model presented here which could be improved by nesting
    704747grids near Aceh.
    705748
     
    708751\includegraphics[width=12.0cm,keepaspectratio=true]{jasonComparison.jpg}
    709752\caption{Comparison of the \textsc{ursga}-predicted surface elevation
    710   with the JASON satellite altimetry data. The \textsc{ursga} wave
     753  with the \textsc{jason} satellite altimetry data. The \textsc{ursga} wave
    711754  heights have been corrected for the time the satellite passed
    712   overhead compared to JASON sea level anomaly.}
     755  overhead compared to \textsc{jason} sea level anomaly.}
    713756\label{fig:jasonComparison}
    714757\end{center}
    715758\end{figure}
     759FIXME (Jane): This graph does not look nice. The legend URS Model should
     760be URSGA model.
    716761
    717762\subsection{Inundation}
    718763After propagating the tsunami in the open ocean using \textsc{ursga},
    719764the approximated ocean and surface elevation and horisontal flow
    720 velocities were extracted and used to construct a boundary condition
     765velocities were extracted and used to construct a boundary condition 
    721766for the \textsc{anuga} model. The interface between the \textsc{ursga}
    722 and \textsc{anuga} models was chosen to roughly follow the 100 m depth
     767and \textsc{anuga} models was chosen to roughly follow the 100~m depth
    723768contour along the west coast of Phuket Island. The computational
    724769domain is shown in Figure~\ref{fig:computational_domain}.
    725 \begin{figure}[ht]
    726 \begin{center}
    727 %\includegraphics[width=5.0cm,keepaspectratio=true]{extent_of_ursga_model.jpg}
    728 %\includegraphics[width=5.0cm,keepaspectratio=true]{ursgaDomain.jpg}
    729 \includegraphics[width=5.0cm,keepaspectratio=true]{extent_of_ANUGA_model.jpg}
    730 \caption{Computational domain of the URSGA simulation (inset: white and black squares and main: black square) and the \textsc{anuga} simulation (main and inset: red polygon).}
    731 \label{fig:computational_domain}
    732 \end{center}
    733 \end{figure}
    734770
    735771The domain was discretised into 386,338 triangles. The resolution of
    736 the grid was increased in certain regions to efficiently increase the
    737 accuracy of the simulation. The grid resolution ranged between a
    738 maximum triangle area of $1\times 10^5$ m$^2$ near the Western ocean
     772the grid was increased in regions inside the bay and on-shore to
     773efficiently increase the simulation accuracy for the impact area.
     774The grid resolution ranged between a
     775maximum triangle area of $1\times 10^5$ m$^2$ near the western ocean
    739776boundary to $20$ m$^2$ in the small regions surrounding the inundation
    740777region in Patong Bay. Due to a lack of available data, friction was
    741778set to a constant throughout the computational domain. For the
    742 reference simulation a Manning's coefficient of 0.01 was chosen to
     779reference simulation, a Manning's coefficient of 0.01 was chosen to
    743780represent a small resistance to the water flow. See Section
    744781\ref{sec:friction sensitivity} for details on model sensitivity to
     
    747784
    748785The boundary condition at each side of the domain towards the south
    749 and the north where no data was available was chosen as a transmissive
     786and the north where no information about the incident wave or
     787its velocity field is available
     788was chosen as a transmissive
    750789boundary condition, effectively replicating the time dependent wave
    751 height present just inside the computational domain. Momentum was set
     790height present just inside the computational domain.
     791The velocity field on these boundaries was set
    752792to zero. Other choices include applying the mean tide value as a
    753 Dirichlet type boundary condition. But experiments as well as the
     793Dirichlet boundary condition. But experiments as well as the
    754794result of the verification reported here showed that this approach
    755795tends to underestimate the tsunami impact due to the tempering of the
     
    761801specified by the Thai Navy tide charts
    762802(\url{http://www.navy.mi.th/hydro/}) at the time the tsunami arrived
    763 at Patong Bay. Although the tsunami propagated for approximately 3
     803at Patong Bay. Although the tsunami propagated for approximately three
    764804hours before it reach Patong Bay, the period of time during which the
    765805wave propagated through the \textsc{anuga} domain is much
     
    767807reasonable.
    768808
    769 Maximum onshore inundation elevation was computed from the model
     809Maximum onshore inundation depth was computed from the model
    770810throughout the entire Patong Bay region.
    771811Figure~\ref{fig:inundationcomparison1cm} (left) shows very good
     
    790830
    791831
     832The datasets necessary for reproducing the results
     833of the inundation stage are available on Sourceforge under the \textsc{anuga}
     834project (\url{http://sourceforge.net/projects/anuga}).
     835At the time of
     836writing the direct link is \url{http://tinyurl.com/patong2004-data}.
     837%%\url{http://sourceforge.net/project/showfiles.php?group_id=172848&package_id=319323&release_id=677531}.
     838The scripts required are part of the \textsc{anuga} distribution also
     839available from Sourceforge \url{http://sourceforge.net/projects/anuga} under
     840the validation section.
     841
    792842An animation of this simulation is available on the \textsc{anuga} website at \url{https://datamining.anu.edu.au/anuga} or directly from \url{http://tinyurl.com/patong2004}.
    793 
    794843%\url{https://datamining.anu.edu.au/anuga/attachment/wiki/AnugaPublications/patong_2004_indian_ocean_tsunami_ANUGA_animation.mov}.
    795844
     
    826875parameterisation of the source model, effect of humans structures on
    827876flow, as well as uncertainties in the elevation data, effects of
    828 erosion and deposition by the tsunami event, measurement errors, and
     877erosion and deposition by the tsunami event,
     878measurement errors in the GPS survey recordings, and
    829879missing data in the field survey data itself. The impact of some of
    830880these sources of uncertainties are is investigated in
     
    852902\includegraphics[width=10.0cm,keepaspectratio=true]{gauge_bay_depth.jpg}
    853903\includegraphics[width=10.0cm,keepaspectratio=true]{gauge_bay_speed.jpg}
    854 \caption{Time series obtained from the two offshore locations shown in Figure \protect \ref{fig:gauge_locations}.}
     904\caption{Time series obtained from the two offshore gauge locations,
     9057C and 10C, shown in Figure \protect \ref{fig:gauge_locations}.}
    855906\end{center}
    856907\label{fig:offshore_timeseries}
     
    861912\includegraphics[width=10.0cm,keepaspectratio=true]{gauges_hotels_depths.jpg}
    862913\includegraphics[width=10.0cm,keepaspectratio=true]{gauges_hotels_speed.jpg}
    863 \caption{Time series obtained from the two onshore locations shown in Figure \protect \ref{fig:gauge_locations}.}
     914\caption{Time series obtained from the two onshore locations, North and South,
     915shown in Figure \protect \ref{fig:gauge_locations}.}
    864916\end{center}
    865917\label{fig:onshore_timeseries}
     
    867919
    868920
    869 The estimated max depths and flow rates given in Section
     921The estimated depths and flow rates given in Section
    870922\ref{sec:eyewitness data} are shown together with the modelled depths
    871923and flow rates obtained from the model in Table \ref{tab:depth and
     
    891943\label{tab:depth and flow comparisons}
    892944\end{table}
     945FIXME (Jane): We should perhaps look at average data in area surrounding these points
    893946
    894947%can be estimated with landmarks found in
     
    921974model maximum inundation. The reference model is the one reported in
    922975Figure~\ref{fig:inundationcomparison1cm} (right) with a friction coefficient of 0.01,
    923 buildings included and the boundary condition produced by the URSGA model.
     976buildings included and the boundary condition produced by the
     977\textsc{ursga} model.
    924978
    925979%========================Friction==========================%
    926980\subsection{Friction}
    927981\label{sec:friction sensitivity}
    928 The first study investigated the impact of surface roughness on the
     982The first sensitivity study investigated the impact of surface roughness on the
    929983predicted run-up. According to Schoettle~\cite{schoettle2007}
    930984appropriate values of Manning's coefficient range from 0.007 to 0.03
     
    9491003severity is directly proportional to the boundary waveheight but small
    9501004perturbations in the input wave height of 10 cm appear to have little
    951 effect on the final on-shore run-up. Obviously larger perturbations
    952 will have greater impact. However, this value is generally well
     1005effect on the final inundated area. Obviously larger perturbations
     1006will have greater impact. However, wave heights in the open ocean are
     1007generally well
    9531008predicted by the generation and propagation models such as
    954 \textsc{ursga}. See e.g.\ \cite{thomas2009}.
     1009\textsc{ursga} as demonstrated in Section \ref{sec:resultsPropagation}
     1010and also in \cite{thomas2009}.
    9551011
    9561012
     
    9581014%========================Buildings==========================%
    9591015\subsection{Buildings and Other Structures}
    960 The presence of buildings has the greatest influence on the maximum
    961 on-shore inundation extent. Figure~\ref{fig:sensitivity_nobuildings}
    962 shows the maximum run-up and associated flow speeds in the presence and absence of buildings. It
    963 is apparent that the inundation is much more severe when the presence
    964 of human made structures and buildings are ignored.
     1016The presence or absence of physical buildings in the elevation model was also
     1017investigated.
     1018Figure~\ref{fig:sensitivity_nobuildings}
     1019shows the inundated area and the associated maximum flow speeds
     1020in the presence and absence of buildings. It
     1021is apparent that densely built-up areas act as
     1022dissipators greatly reducing the inundated area. However, flow speeds
     1023tend to increase in passages between buildings.
     1024 
    9651025
    9661026\begin{table}
     
    10011061
    10021062This study also shows that the tsunami impact modelling methodology
    1003 adopted is sane and able to predict inundation extents with reasonable
     1063adopted is credible and able to predict inundation extents with reasonable
    10041064accuracy.  An associated aim of this paper was to further validate the
    10051065hydrodynamic modelling tool \textsc{anuga} which is used to simulate
    1006 the tsunami inundation and run rain-induced floods. Model predictions
    1007 matched well geodetic measurements of the Sumatra--Andaman earthquake,
     1066the tsunami inundation. Model predictions
     1067matched well the geodetic measurements of the Sumatra--Andaman earthquake,
    10081068altimetry data from the \textsc{jason}, eye-witness accounts of wave
    10091069front arrival times and flow speeds and a detailed inundation survey
     
    10131073small changes in friction, wave height at the 100 m depth contour and
    10141074the presence of buildings and other structures on the model
    1015 predictions. The presence of buildings has the greatest influence on
     1075predictions. Of these three, the presence of buildings was shown to
     1076have the greatest influence on
    10161077the simulated inundation extent. The value of friction and small
    10171078perturbations in the waveheight at the \textsc{anuga} boundary have
     
    10221083This project was undertaken at Geoscience Australia and the Department
    10231084of Mathematics, The Australian National University. The authors would
    1024 like to thank Niran Chaimanee from the CCOP, Thailand for providing
     1085like to thank Niran Chaimanee from the CCOP for providing
    10251086the post 2004 tsunami survey data, building footprints, aerial
    1026 photography and the elevation data for Patong beach, Prapasri Asawakun
     1087photography and the elevation data for Patong city, Prapasri Asawakun
    10271088from the Suranaree University of Technology and Parida Kuneepong for
    10281089supporting this work; and Drew Whitehouse from the Australian National
    1029 University for preparing the animation of the inundation model.
     1090University for preparing the animation of the simulated impact.
    10301091
    10311092\clearpage
     
    10411102\caption{Results from reference model as reported in Section \protect \ref{sec:results},
    10421103  i.e.\ including buildings and a friction value of 0.01. The seaward boundary condition is as
    1043   provided by the URSGA model. The left image shows the maximum
     1104  provided by the \textsc{ursga} model. The left image shows the maximum
    10441105  modelled depth while the right hand image shows the maximum modelled
    10451106  flow velocities.}
     
    10581119  \protect \ref{fig:reference_model} (left).  The left and right images
    10591120  show the inundation results if the wave at the \textsc{anuga} boundary
    1060   is reduced or increased by 10cm respectively. The inundation
     1121  is reduced or increased by 10 cm respectively. The inundation
    10611122  severity varies in proportion to the boundary waveheight, but the
    10621123  model results are only slightly sensitive to this parameter for the
     
    10651126\end{center}
    10661127\end{figure}
     1128FIXME (Jane): How and why was the +/- 10 cm chosen?
    10671129
    10681130
     
    10831145\includegraphics[width=6cm,keepaspectratio=true]{sensitivity_nobuildings_depth}
    10841146\includegraphics[width=6cm,keepaspectratio=true]{sensitivity_nobuildings_speed}
    1085 \caption{This figure shows the effect of having buildings as part of
     1147\caption{Model results show the effect of buildings in
    10861148  the elevation data set.
    1087   The left hand image shows the inundation depth results for
     1149  The left hand image shows the maximum inundation depth results for
    10881150  a model entirely without buildings.  As expected, the absence of
    10891151  buildings will increase the inundation extent beyond what was
Note: See TracChangeset for help on using the changeset viewer.