Changeset 7480


Ignore:
Timestamp:
Sep 4, 2009, 1:29:13 PM (15 years ago)
Author:
ole
Message:

Multiple revisions of text.
Moved labels from inside center block as this was causing wrong numbers.

Location:
anuga_work/publications/boxing_day_validation_2008
Files:
8 edited

Legend:

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

    r7467 r7480  
    88\begin{center}
    99\includegraphics[width=\textwidth,keepaspectratio=true]{figures/reference}
     10\end{center}
     11
    1012\caption{Results from reference model as reported in Section
    1113\protect \ref{sec:results},
     
    1618  flow velocities.}
    1719\label{fig:reference_model}
    18 \end{center}
    1920\end{figure}
    2021
     
    2223\begin{center}
    2324\includegraphics[width=\textwidth,keepaspectratio=true]{figures/sensitivity_friction}
     25\end{center}
     26
    2427\caption{Model results for different values of Manning's friction
    2528  coefficient shown to assess sensitivities.
     
    3437  scope of this study.}
    3538\label{fig:sensitivity_friction}
    36 \end{center}
    3739\end{figure}
    3840
     
    4042\begin{center}
    4143\includegraphics[width=\textwidth,keepaspectratio=true]{figures/sensitivity_friction_speed}
     44\end{center}
     45
    4246\caption{The maximal flow speeds for the same model parameterisations
    43   found in Figure \protect \ref{fig:sensitivity_friction}.}
     47  found in Figure \protect \ref{fig:sensitivity_friction}.
     48  The reference flow speeds for a
     49  friction value of 0.01 are shown in Figure
     50  \protect \ref{fig:reference_model} (right).}
    4451\label{fig:sensitivity_friction_speed}
    45 \end{center}
    4652\end{figure}
    4753
     
    5056\begin{center}
    5157\includegraphics[width=\textwidth,keepaspectratio=true]{figures/sensitivity_boundary_wave}
     58\end{center}
     59
    5260\caption{Model results with wave height at \textsc{anuga} boundary artificially
    5361  modified to assess sensitivities.
     
    5563  \protect \ref{fig:reference_model} (left). The left and right images
    5664  show the inundation results if the wave at the \textsc{anuga} boundary
    57   is reduced or increased by 10 cm respectively. The inundation
     65  is increased or reduced by 10 cm respectively. The inundation
    5866  severity varies in proportion to the boundary waveheight, but the
    5967  model results are only slightly sensitive to this parameter for the
    6068  range of values tested.}
    6169\label{fig:sensitivity_boundary}
    62 \end{center}
    6370\end{figure}
    6471
     
    6774\begin{center}
    6875\includegraphics[width=\textwidth,keepaspectratio=true]{figures/sensitivity_boundary_wave_speed}
     76\end{center}
     77
    6978\caption{The maximal flow speeds for the same model parameterisations
    70   found in Figure \protect \ref{fig:sensitivity_boundary}.}
     79  found in Figure \protect \ref{fig:sensitivity_boundary}.
     80  The reference flow speeds are shown in Figure
     81  \protect \ref{fig:reference_model} (right).}
    7182\label{fig:sensitivity_boundary_speed}
    72 \end{center}
    7383\end{figure}
    7484
     
    7787\begin{center}
    7888\includegraphics[width=\textwidth,keepaspectratio=true]{figures/sensitivity_buildings}
     89\end{center}
    7990\caption{Model results show the effect of buildings in
    8091  the elevation data set.
    8192  The left hand image shows the inundation extent as modelled in the reference
    82   model (Figure \protect \ref{fig:reference_model}) which includes
     93  model (Figure \protect \ref{fig:reference_model}, left) which includes
    8394  buildings in the elevation data. The right hand image
    8495  shows the result for a bare-earth model i.e. entirely without buildings. 
     
    8798  surveyed.}
    8899\label{fig:sensitivity_buildings}
    89 \end{center}
    90100\end{figure}
    91101
     
    94104\begin{center}
    95105\includegraphics[width=\textwidth,keepaspectratio=true]{figures/sensitivity_buildings_speed}
     106\end{center}
    96107\caption{The maximal flow speeds for the same model parameterisations
    97108  found in Figure~\protect~\ref{fig:sensitivity_buildings}.
     
    99110  but tends to increase speeds in passages between buildings.}
    100111\label{fig:sensitivity_buildings_speed}
    101 \end{center}
    102112\end{figure}
  • anuga_work/publications/boxing_day_validation_2008/conclusion.tex

    r7451 r7480  
    11\section{Conclusion}
    2 This paper proposes an additional field data benchmark for the
     2This paper proposes a new field data benchmark for the
    33verification of tsunami inundation models. Currently, there is a
    44scarcity of appropriate validation datasets due to a lack of well-documented
     
    1818Australiawhich is used to simulate the tsunami inundation.
    1919This study ashows that the tsunami  modelling methodology adopted is credible
    20 and able to predict detailed inundation extents with reasonable accuracy.
     20and able to predict detailed inundation extents and dynamics with reasonable accuracy.
    2121Model predictions matched well a detailed inundation survey
    2222of Patong Bay, Thailand as well as altimetry data from the \textsc{jason},
     
    2525A simple sensitivity analysis was performed to assess the influence of
    2626small changes in friction, wave height at the 100 m depth contour and
    27 the presence of buildings and other structures on the model
     27the presence of buildings on the model
    2828predictions. Of these three, the presence of buildings was shown to
    2929have the greatest influence on the simulated inundation extent. This result
    3030indicates that the influence of human-made structures should be included,
    31 where possible in any future studies. The value of friction and small
     31where possible, in any future studies. The value of friction and small
    3232perturbations in the waveheight at the \textsc{anuga} boundary have
    3333comparatively little effect on the model results.
  • anuga_work/publications/boxing_day_validation_2008/data.tex

    r7473 r7480  
    1818However, for physics-based models evaluation of the model during the generation
    1919 and propagation phases is still useful. If a model is physics-based one
    20 should ensure that all physics are being modelled accurately. Moreover
     20should ensure that all physics are being modelled accurately. Moreover,
    2121evaluation of all three stages of tsunami evolution can help identify the
    2222cause of any discrepancies between modelled and observed inundation.
    23 Consequently in this section we present data not only to facilitate
     23Consequently, in this section we present data not only to facilitate
    2424validation of inundation but to also aid the assessment of tsunami
    2525generation and propagation.
     
    9090\item a one second grid created from the digitised Thai Navy
    9191  bathymetry chart, no. 358, which covers Patong Bay and the
    92   immediately adjacent regions.
    93   The gridding of data was performed using {\bf Intrepid}, a commercial
     92  immediately adjacent regions. The digitised points and contour lines
     93from this chart are shown in Figure~\ref{fig:patong_bathymetry}.
     94  The gridding of data was performed using \textsc{Intrepid}, a commercial
    9495  geophysical processing package developed by Intrepid Geophysics. The
    9596  gridding scheme employed the nearest neighbour algorithm followed by
     
    113114A subset of the nine second grid was replaced by the three second
    114115data. Finally, the one second grid was used to approximate the
    115 bathymetry in Patong Bay and the immediately adjacent regions. Any
     116bathymetry in Patong Bay. Any
    116117points that deviated from the general trend near the boundary were
    117118deleted as a quality check.
     
    119120A one second grid was used to approximate the bathymetry in Patong
    120121Bay. This elevation data was created from the digitised Thai
    121 Navy bathymetry chart, no 358. The digitised points and contour lines
    122 from this chart are shown in Figure~\ref{fig:patong_bathymetry}.
    123 
    124 
    125 The sub-sampling of larger grids was performed by using {\bf resample},
     122Navy bathymetry chart, no 358.
     123
     124
     125The sub-sampling of larger grids was performed by using \textsc{resample},
    126126a Generic Mapping Tools (\textsc{GMT}) program (\cite{wessel98}).
    127127
     
    130130\begin{center}
    131131\includegraphics[width=\textwidth,keepaspectratio=true]{figures/nested_grids}
     132\end{center}
     133
    132134\caption{Nested bathymetry grids.}
    133135\label{fig:nested_grids}
    134 \end{center}
    135136\end{figure}
    136137
     
    149150\subsection{Inundation}
    150151\label{sec:inundation data}
    151 Inundation is the final stage of the evolution of a tsunami and
    152 refers to the run-up of tsunami onto land. This process is typically the most
     152Inundation is the final stage of the evolution of a tsunami and refers
     153to the run-up of tsunami onto land. This process is typically the most
    153154difficult of the three stages to model due to thin layers of water
    154155flowing rapidly over dry land.  Aside from requiring robust solvers
     
    157158data which is often not available. In the case of model validation
    158159high quality field measurements are also required. For the proposed
    159 benchmark a high resolution (1 second) topography data set and a
    160 tsunami inundation survey map from the
    161 Coordinating Committee Co-ordinating Committee for Geoscience Programmes
    162 in East and Southeast Asia (CCOP) (\cite{szczucinski06}) was obtained
    163 to validate model inundation. See also acknowledgements at the end of this paper. In this section we also present eye-witness accounts which can be used
    164 to qualitatively validate tsunami inundation.
     160benchmark a high resolution topography data set (in the form of GIS
     161contour lines) and a tsunami inundation survey map from the
     162Coordinating Committee Co-ordinating Committee for Geoscience
     163Programmes in East and Southeast Asia (CCOP) (\cite{szczucinski06})
     164was obtained to validate model inundation. See also acknowledgements
     165at the end of this paper. In this section we also present eye-witness
     166accounts which can be used to qualitatively validate tsunami
     167inundation.
    165168
    166169\subsubsection{Topography Data}
    167 A 1 second grid comprising the onshore topography and the nearshore bathymetry
    168 for Patong Beach was created from the Navy charts (described in Section \ref{sec:bathymetry data}) and from
    169 1 m and 10 m elevation contours provided in a GIS data set which was also provided by the CCOP
    170 (see Section \ref{sec:inundation data} for details).
    171 The 1 second terrain model for the and community as shown in Figure~\ref{fig:patong_bathymetry}.
    172 
    173 Two 1/3 second grids were created: One for the saddle point covering Merlin and Tri Trang Beaches
    174 and one for Patong city and its immediate shore area.
    175 These grids were based on
    176 the same data used for the 1 second data grid.
    177 The patong city grid was further modified based on
    178 satellite imagery to include
    179 the river and lakes towards the south of Patong city which were not part of the GIS dataset.
    180 The depth of the river and lake system was set uniformly to -1 m.
     170A 1 second grid comprising the onshore topography and the nearshore
     171bathymetry for Patong Beach was created from the Navy charts
     172(described in Section \ref{sec:bathymetry data}) and from 1 m and 10 m
     173elevation contours provided by the CCOP (see Section
     174\ref{sec:inundation data} for details). The 1 second terrain model
     175for the and community as shown in Figure~\ref{fig:patong_bathymetry}.
     176
     177Two 1/3 second grids were created: One for the saddle point covering
     178Merlin and Tri Trang Beaches and one for Patong City and its immediate
     179shore area.  These grids were based on the same data used for the 1
     180second data grid.  The Patong city grid was further modified based on
     181satellite imagery to include the river and lakes towards the south of
     182Patong City which were not part of the provided elevation data.
     183The depth of the river and lake system was set uniformly to a depth of 1 m.
    181184
    182185
     
    184187\begin{center}
    185188\includegraphics[width=8.0cm,keepaspectratio=true]{figures/patong_bay_data.jpg}
     189\end{center}
     190
    186191\caption{3D visualisation of the elevation data set used for the nearshore propagation and and inundation in Patong Bay showing
    187192digitised data points and contours as well as rivers and roads
    188193draped over the data model.}
    189194\label{fig:patong_bathymetry}
    190 \end{center}
    191195\end{figure}
    192196
     
    195199Human-made buildings and structures can significantly affect tsunami
    196200inundation. The footprint and number of floors of the
    197 buildings in Patong Bay were extracted from the GIS data set from CCOP.
     201buildings in Patong Bay were extracted from the data provided by CCOP.
    198202The heights of these
    199203buildings were estimated assuming that each floor has a height of 3 m and they
     
    219223\begin{center}
    220224\includegraphics[width=\textwidth,keepaspectratio=true]{figures/post_tsunami_survey.jpg}
     225\end{center}
     226
    221227\caption{Tsunami survey mapping the maximum observed inundation at
    222228  Patong beach courtesy of the CCOP \protect \cite{szczucinski06}.}
    223229\label{fig:patongescapemap}
    224 \end{center}
    225230\end{figure}
    226231
     
    240245\begin{center}
    241246\includegraphics[width=\textwidth,keepaspectratio=true]{figures/gauges.jpg}
     247\end{center}
     248
    242249\caption{Location of timeseries extracted from the model output.}
    243 % FIXME(John):
    244 %should we combine the inundation map with the gauages map?}
    245 %OLE: No - Kristy and I tried to make each image very simple -
    246 %with one message only
    247250\label{fig:gauge_locations}
    248 \end{center}
    249251\end{figure}
    250252
     
    275277\includegraphics[width=5.0cm,keepaspectratio=true]{figures/flow_rate_south_7_12sec.jpg}
    276278\includegraphics[width=5.0cm,keepaspectratio=true]{figures/flow_rate_south_7_60sec.jpg}
     279\end{center}
     280
    277281\caption{Four frames from a video where flow rate could be estimated,
    278   circle indicates tracked debris, from top left: 0.0 sec, 5.0 s, 7.1
     282  circle indicates tracked debris, from top left: 0.0 s, 5.0 s, 7.1
    279283  s, 7.6 s.}
    280284\label{fig:video_flow}
    281 \end{center}
    282285\end{figure}
    283286
    284287Flow rates were estimated using landmarks found in both videos and
    285 were found to be in the range of 5 to 7 metres per second (+/- 2 m/s)
    286 in the north and 0.5 to 2 metres per second (+/- 1 m/s) in the
     288were found to be in the range of 5 to 7 m/s ($\pm$2 m/s)
     289in the north and 0.5 to 2 m/s ($\pm$1 m/s) in the
    287290south\footnote{These error bounds were estimated from uncertainty in aligning the debris with building boundaries in the videos.}.
    288291Water depths could also
    289292be estimated from the videos by the level at which water rose up the
    290293sides of buildings such as shops. Our estimates are in the order of
    291 1.5 to 2.0 metres (+/- 0.5 m estimated error bounds).
    292 Fritz ~\cite{fritz06} performed a detailed
     2941.5 to 2.0 m ($\pm$0.5 m estimated error bounds).
     295Fritz~\cite{fritz06} performed a detailed
    293296analysis of video frames taken around Banda Aceh and arrived at flow
    294297speeds in the range of 2 to 5 m/s.
  • anuga_work/publications/boxing_day_validation_2008/introduction.tex

    r7451 r7480  
    1919non-linear three-dimensional mechanical models~\cite{zhang08}.
    2020These models are typically used to predict quantities such as arrival
    21 times, wave speeds and heights, and inundation extents
    22 which are used to develop efficient hazard mitigation plans. Physics based
     21times, wave speeds and heights, as well as inundation extents
     22which can be used to develop efficient hazard mitigation plans. Physics based
    2323models combine observed seismic, geodetic and sometimes tsunami data to
    2424 provide estimates of initial sea floor and ocean surface
    25 deformation. Whilst the shallow water wave equations~\cite{george06},
     25deformation. The shallow water wave equations~\cite{george06},
    2626linearised shallow water wave equations~\cite{liu09},
    2727and Boussinesq-type equations~\cite{weiss06} are frequently used to simulate
     
    3737established.  One can only hope to state under what conditions and to
    3838what extent the
    39 model hypothesis holds true. Specifically the utility of a model can
     39model hypothesis holds true. Specifically, the utility of a model can
    4040be assessed through a process of verification and
    4141validation. Verification assesses the accuracy of the numerical method
     
    114114 tailored accordingly.
    115115
    116 Unlike the existing field benchmarks the proposed test does facilitate
     116Unlike the existing field benchmarks the proposed test facilitates
    117117 localised and highly detailed spatially distributed assessment of
    118118modelled inundation. To the authors knowledge it is also the first benchmark to
    119 assess model inundation under influenced by numerous human structures.
    120 Eye-witness videos also allow the qualitative assessment of onshore flow
     119assess model inundation influenced by numerous human structures.
     120Eye-witness videos have also been considered to allow the qualitative assessment of onshore flow
    121121patterns.
    122122
  • anuga_work/publications/boxing_day_validation_2008/method.tex

    r7470 r7480  
    33generation, propagation and run-up~\cite{titov97a,satake95}. Here we
    44introduce the modelling methodology employed by Geoscience Australia
    5 to illustrate the utility of the proposed benchmark. The methodology used by Geoscience Australia has three distinct components. Firstly an appropriate model is used to approximate the initial sea surface deformation. This model is chosen according to the cause of the intial distrubance. The resulting wave is propagated using \textsc{ursga} in the deep ocean until the wave reaches shallow water, typically the $100$m depth contour. The ocean surface profile along this contour is used as a time varying boundary condition for ANUGA which simulates the propagation of the tsunami within the shallow water and the subsequent inundation of the land. This three part methodology roughly follows the three stages of tsunami evolution. The components used to model each stage of evolution are described in more detail below.
     5to illustrate the utility of the proposed benchmark. The methodology used by Geoscience Australia has three distinct components. Firstly an appropriate model is used to approximate the initial sea surface deformation. This model is chosen according to the cause of the intial disturbance. The resulting wave is propagated using the \textsc{ursga} model (see Section~\ref{sec:ursga}) in the deep ocean until the wave reaches shallow water, typically the $100$ m depth contour. The ocean surface profile along this contour is used as a time varying boundary condition for the \textsc{anuga} model (see Section~\ref{sec:anuga}) which simulates the propagation of the tsunami within the shallow water and the subsequent inundation of the land. This three part methodology roughly follows the three stages of tsunami evolution. The components used to model each stage of evolution are described in more detail below.
    66
    77\subsection{Generation}\label{sec:modelGeneration}
     
    5757et al~\cite{chlieh07}. This model was created by inversion of wide
    5858range of geodetic and seismic data. The slip model consists of 686
    59 20km x 20km subsegments each with a different slip, strike and dip
     5920 km x 20 km subsegments each with a different slip, strike and dip
    6060angle. The dip subfaults go from $17.5^0$ in the north and $12^0$ in
    6161the south. Refer to Chlieh et al~\cite{chlieh07} for a detailed
     
    8181%The propagation of the tsunami in shallow water ($<100$m) and inundation are modelled using a hydrodynamic package called \textsc{ursga}. This package is ideally suited to shallow water propagation and inundation as it accurately simulates flow over dry land and is based upon an irregular triangular grid which can be refined in areas of interest.
    8282
    83 \subsubsection{URSGA}
     83\subsubsection{URSGA}\label{sec:ursga}
    8484\textsc{ursga} is a hydrodynamic code that models the propagation of
    8585the tsunami in deep water using a finite difference method on a staggered grid.
     
    115115% A description of \textsc{anuga} is the following section.
    116116
    117 \subsubsection{ANUGA}
     117\subsubsection{ANUGA}\label{sec:anuga}
    118118\textsc{Anuga} is a Free and Open Source hydrodynamic inundation tool that
    119119solves the conserved form of the depth-integrated nonlinear shallow
     
    133133are allowed at every edge in the mesh. This means that the tool is
    134134well suited to adequately resolving hydraulic jumps, transcritical flows and
    135 the process of wetting and drying. This means that \textsc{Anuga}
     135the process of wetting and drying. Consequently, \textsc{anuga}
    136136is suitable for
    137137simulating water flow onto a beach or dry land and around structures
    138 such as buildings. \textsc{Anuga} has been validated against
     138such as buildings. \textsc{anuga} has been validated against
    139139%a number of analytical solutions !!!Analytical solutions have not been published. Ask Steve.
    140140the wave tank simulation of the 1993 Okushiri
    141141Island tsunami~\cite{nielsen05,roberts06} and
    142142dam break experiments~\cite{baldock07}.
     143More information on \textsc{anuga} and how to obtain it are available from \url{https://datamining.anu.edu.au/anuga}.
     144
  • anuga_work/publications/boxing_day_validation_2008/paper.tex

    r7467 r7480  
    4444which affords a uniquely large amount of observational data for
    4545events of this kind. The proposed benchmark is intended to aid
    46 validation of tsunami inundation, which is the most important stage
    47 of tsunami evolution. However individual tests are presented to
     46validation of tsunami inundation modelling, which is the most important stage
     47of tsunami evolution. However, individual tests are presented to
    4848facilitate model evaluation for the generation and propagation
    49 phases as well. Specifically we use geodetic
     49phases as well. Specifically, we use geodetic
    5050measurements of the Sumatra--Andaman earthquake to validate the
    5151tsunami source, altimetry data from the \textsc{jason} satellite to
  • anuga_work/publications/boxing_day_validation_2008/results.tex

    r7474 r7480  
    22This section presents a validation of the modelling practice of Geoscience
    33Australia against the new proposed benchmarks. The criteria outlined
    4 in Section~\ref{sec:checkList} are addressed.S
     4in Section~\ref{sec:checkList} are addressed.
    55
    66\subsection{Generation}\label{modelGeneration}
     
    3434not surprising, since the original slip model was chosen
    3535by~\cite{chlieh07} to fit the motion and seismic data well.
    36 Consequently the replication of the generation data has little meaning for
     36Consequently, the replication of the generation data has limited meaning for
    3737the model structure presented in Section~\ref{sec:models}. But for
    3838uncalibrated source models or source models calibrated on other data
     
    4242\begin{center}
    4343\includegraphics[width=0.8\textwidth,keepaspectratio=true]{figures/surface_deformation.jpg}
     44\end{center}
     45
    4446\caption{Location and magnitude of the vertical component of the sea
    4547  floor displacement associated with the 2004 Indian Ocean tsunami
     
    5759  by~\cite{chlieh07}.}
    5860\label{fig:surface_deformation}
    59 \end{center}
    6061\end{figure}
    6162
     
    7677\begin{center}
    7778\includegraphics[width=0.8\textwidth,keepaspectratio=true]{figures/extent_of_ANUGA_model.jpg}
     79\end{center}
     80
    7881\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).}
    7982\label{fig:computational_domain}
    80 \end{center}
    8183\end{figure}
    8284
     
    104106\begin{center}
    105107\includegraphics[width=\textwidth,keepaspectratio=true]{figures/jasonComparison.jpg}
     108\end{center}
     109
    106110\caption{Comparison of the \textsc{ursga}-predicted surface elevation
    107111  with the \textsc{jason} satellite altimetry data. The \textsc{ursga} wave
     
    109113  overhead compared to \textsc{jason} sea level anomaly.}
    110114\label{fig:jasonComparison}
    111 \end{center}
    112 \end{figure}
    113 FIXME (Jane): This graph does not look nice.
     115\end{figure}
    114116
    115117After propagating the tsunami in the open ocean using \textsc{ursga},
     
    133135approximately 6 m between mesh points)
    134136in the small regions surrounding the inundation
    135 region in Patong Bay. The coarse resolution was chosen to be
    136 commensurate with the model output from the \textsc{ursga} model (FIXME (Ole): Richard says that the ursga model used all four grids which would mean that the
    137 resolution at the ANUGA boundary was 1 second or about 30 m.
    138 This is not consistent with my memory and certainly not with us choosing a
    139 resolution of 440 m. John, do you remember what the spacing was between the
    140 URSGA points? Did we weed them out or did we take them as they were?)
    141 
    142 while the latter was chosen to match the available resolution of topographic
    143 data and building data in Patong city.
     137region in Patong Bay. The coarse resolution was chosen to balance accuracy
     138with computational costs while the fine resolution was chosen to match the available resolution of topographic
     139data and building data in Patong City. Figure~\ref{fig:mesh} shows a
     140section of the mesh covering the southern part of the City.
     141
     142\begin{figure}[ht]
     143\begin{center}
     144\includegraphics[width=0.8\textwidth, keepaspectratio=true]{figures/mesh_section.jpg}
     145\end{center}
     146
     147\caption{Section of the mesh used by ANUGA to simulate the tsunami inundation.
     148The finest mesh resolution is approximately 6 m between nodes which
     149is sufficient to resolve individual buildings affecting the flows.}
     150\label{fig:mesh}
     151\end{figure}
     152
     153
    144154Due to a lack of available roughness data, friction was
    145155set to a constant throughout the computational domain. For the
     
    169179(\url{http://www.navy.mi.th/hydro/}) at the time the tsunami arrived
    170180at Patong Bay. Although the tsunami propagated for approximately three
    171 hours before it reach Patong Bay, the period of time during which the
     181hours before it reached Patong Bay, the period of time during which the
    172182wave propagated through the \textsc{anuga} domain is much
    173 smaller. Consequently the assumption of constant tide height is
     183smaller of the order of 2 hours. Consequently the assumption of constant tide height is
    174184reasonable.
    175185
    176186\subsection{Inundation}\label{sec:inundation results}
    177187The \textsc{anuga} simulation described in the previous section and used to
    178  model shallow water propgation also predicts
     188 model shallow water propagation also predicts
    179189inundation. Maximum onshore inundation depth was computed from the model
    180190throughout the entire Patong Bay region and used to generate
    181191a measure of the inundated area.
    182192Figure~\ref{fig:inundationcomparison1cm} (left) shows very good
    183 agreement between the measured and simulated inundation. However
     193agreement between the measured and simulated inundation. However,
    184194these results are dependent on the classification used to determine
    185195whether a region in the numerical simulation was inundated. In
    186196Figure~\ref{fig:inundationcomparison1cm} (left) a point in the computational
    187197domain was deemed inundated if at some point in time it was covered by
    188 at least 1 cm of water. However, the precision of the inundation boundary
     198at least 1 cm of water. The precision of the inundation boundary
    189199generated by the on-site survey is most likely less than that as it
    190200was determined by observing water marks and other signs
    191 left by the receding waters. Consequently the measurement error along
     201left by the receding waters. Consequently, the measurement error along
    192202the inundation boundary of the survey is likely to vary significantly
    193203and somewhat unpredictably.
     
    198208area.
    199209Figure~\ref{fig:inundationcomparison1cm} (right) shows the simulated
    200 inundation using a larger threshold of 10 cm.
    201 
    202 
    203 The datasets necessary for reproducing the results
    204 of the inundation stage are available on Sourceforge under the \textsc{anuga}
    205 project (\url{http://sourceforge.net/projects/anuga}).
    206 At the time of
    207 writing the direct link is \url{http://tinyurl.com/patong2004-data}.
    208 %%\url{http://sourceforge.net/project/showfiles.php?group_id=172848&package_id=319323&release_id=677531}.
    209 The scripts required are part of the \textsc{anuga} distribution also
    210 available from Sourceforge \url{http://sourceforge.net/projects/anuga} under
    211 the validation section.
    212 
     210inundation using a larger threshold of 10 cm and Figure~\ref{fig:anuga screenshot} shows a screenshot from the inundation model.
    213211An 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}.
    214212%\url{https://datamining.anu.edu.au/anuga/attachment/wiki/AnugaPublications/patong_2004_indian_ocean_tsunami_ANUGA_animation.mov}.
     
    217215\begin{center}
    218216\includegraphics[width=\textwidth,keepaspectratio=true]{figures/threshold.jpg}
     217\end{center}
     218
    219219\caption{Simulated inundation versus observed inundation using an
    220220inundation threshold of 1 cm (left) and 10 cm (right).}
    221221\label{fig:inundationcomparison1cm}
    222 \end{center}
    223 \end{figure}
     222\end{figure}
     223
     224\begin{figure}[ht]
     225\begin{center}
     226\includegraphics[width=0.8\textwidth,keepaspectratio=true]{figures/anuga_screenshot.jpg}
     227\end{center}
     228
     229\caption{Screenshot of the \textsc{anuga} simulation of the inundation
     230 at Patong City at local time 10:53 about one hour after the arrival of
     231 the first wave.}
     232\label{fig:anuga screenshot}
     233\end{figure}
     234
     235
     236\subsubsection{Comparison to survey}
    224237
    225238To quantify the agreement between the observed and simulated inundation we
     
    244257Discrepancies between the survey data and the modelled inundation
    245258include: unknown distribution of surface roughness, inappropriate
    246 parameterisation of the source model, effect of humans structures on
    247 flow, as well as uncertainties in the elevation data, effects of
     259parameterisation of the source model, discretisation errors,
     260effect of humans structures on
     261flow, as well as uncertainties in the elevation data, friction, effects of
    248262erosion and deposition by the tsunami event,
    249263measurement errors in the GPS survey recordings, and
     
    252266Section~\ref{sec:sensitivity}.
    253267
     268
     269\subsubsection{Reproducibility of inundation results}
     270As one aim of this paper is to provide a new benchmark for tsunami
     271inundation modelling we have made the datasets available
     272available on \textsc{Sourceforge} in \textsc{anuga}
     273project (\url{http://sourceforge.net/projects/anuga}) under the directory
     274\url{validation\_data/patong-1.0}.
     275At the time of
     276writing a direct link is \url{http://tinyurl.com/patong2004-data}.
     277%%\url{http://sourceforge.net/project/showfiles.php?group_id=172848&package_id=319323&release_id=677531}.
     278
     279To reproduce the inundation modelling results using this data, the reader
     280will need to run the validation scripts (\url{anuga\_validation/automated\_validation\_tests/patong\_beach\_validation}) which are part of the
     281\textsc{anuga} distribution also available from
     282Sourceforge \url{http://sourceforge.net/projects/anuga}.
     283
     284
    254285\subsection{Eye-witness accounts}
    255286
    256287\subsubsection{Arrival time}
    257288The arrival time of the first wave took place between 9:55 and 10:55 as described in
    258 Section~\ref{sec:eyewitness data}. The modelled arrival time at the beach is 10:01
     289Section~\ref{sec:eyewitness data}. The modelled arrival time at the beach is 10:00
    259290as can be verified from the animation provided in
    260291Section \ref{sec:inundation results}.
    261292Subsequent waves of variable magnitude appear over the next two hours
    262293approximately 20-30 minutes apart.
    263 % 10:01, 10:19, 10:46, 11:13, 11:43
    264 The first arrival and overall dynamic behaviour is therefor reasonably consistent with the
     294% 10:00, 10:18, 10:45, 11:12, 11:42
     295The first arrival and overall dynamic behaviour is therefore reasonably consistent with the
    265296eye-witness accounts.
    266297
    267298\subsubsection{Observed wave dynamics}
    268299Figure \ref{fig:gauge_locations} shows four locations where time
    269 series have been extracted from the model. The two offshore time series
     300series have been extracted from the model.
     301The two offshore time series
    270302are shown in Figure \ref{fig:offshore_timeseries} and the two onshore
    271303timeseries are shown in Figure \ref{fig:onshore_timeseries}. The
     
    275307\begin{figure}[ht]
    276308\begin{center}
    277 \includegraphics[width=0.8\textwidth,keepaspectratio=true]{figures/gauge_bay_depth.jpg}
    278 \includegraphics[width=0.8\textwidth,keepaspectratio=true]{figures/gauge_bay_speed.jpg}
     309\includegraphics[width=0.7\textwidth,keepaspectratio=true]{figures/gauge_bay_depth.jpg}
     310\includegraphics[width=0.7\textwidth,keepaspectratio=true]{figures/gauge_bay_speed.jpg}
     311\end{center}
     312
    279313\caption{Time series obtained from the two offshore gauge locations,
    2803147C and 10C, shown in Figure \protect \ref{fig:gauge_locations}.}
    281 \end{center}
    282315\label{fig:offshore_timeseries}
    283316\end{figure}
     
    285318\begin{figure}[ht]
    286319\begin{center}
    287 \includegraphics[width=\textwidth,keepaspectratio=true]{figures/gauges_hotels_depths.jpg}
    288 \includegraphics[width=\textwidth,keepaspectratio=true]{figures/gauges_hotels_speed.jpg}
     320\includegraphics[width=0.7\textwidth,keepaspectratio=true]{figures/gauges_hotels_depths.jpg}
     321\includegraphics[width=0.7\textwidth,keepaspectratio=true]{figures/gauges_hotels_speed.jpg}
     322\end{center}
     323
    289324\caption{Time series obtained from the two onshore locations, North and South,
    290325shown in Figure \protect \ref{fig:gauge_locations}.}
    291 \end{center}
    292326\label{fig:onshore_timeseries}
    293327\end{figure}
     
    319353  \end{array}
    320354\]
     355\caption{Observed depth and flows from the video footage compared to values extracted from the inundation model.}
    321356\label{tab:depth and flow comparisons}
    322357\end{table}
     
    339374%as seen in the figure.
    340375
     376
     377
     378
     379
  • anuga_work/publications/boxing_day_validation_2008/sensitivity.tex

    r7467 r7480  
    99However, model uncertainty should not be ignored. The aim of this section is
    1010not to provide a detailed investigation of sensitivity but to rather
    11 illustrate that changes in important parameters of the \textsc{usrga--anuga}
    12 model  produce behaviour consistent with the known physics and that
     11illustrate that changes in important parameters produce behaviour consistent with the known physics and that
    1312small changes in these parameters produce bounded variations in the output.
    1413
     
    1615friction coefficient, changing waveheight at the 100 m depth contour,
    1716and the presence and absence of buildings in the elevation dataset on
    18 model maximum inundation.
     17model maximum inundation as computed by \textsc{anuga}.
    1918
    2019The reference model is the one reported in
     
    3231we simulated the maximum onshore inundation using a Manning's
    3332coefficient of 0.0003 and 0.03. The resulting inundation maps are
    34 shown in Figure~\ref{fig:sensitivity_friction}
    35 % and the maximum flow speeds in Figure~\ref{fig:sensitivity_friction_speed}.
    36  The figure, along with Table~\ref{table:inundationAreas},
     33shown in Figure~\ref{fig:sensitivity_friction} 
     34and the maximum flow speeds in Figure~\ref{fig:sensitivity_friction_speed}.
     35The figure, along with Table~\ref{table:inundationAreas},
    3736shows that the on-shore inundation extent decreases with increasing
    3837friction and that small perturbations in the friction cause bounded
     
    5352than the expected error in the amplitude predicted by the propagation model.
    5453
    55 Figure~\ref{fig:sensitivity_boundary} and  Table~\ref{table:inundationAreas}
     54Figure~\ref{fig:sensitivity_boundary}, Figure~\ref{fig:sensitivity_boundary_speed},
     55and  Table~\ref{table:inundationAreas}
    5656indicate that the inundation severity is directly proportional to the
    5757boundary waveheight but small
     
    7777human-made structures should be included into the model topography.
    7878Furthermore, these results also
    79 indicate that simply matching point sites with much lower resolution meshes
     79indicate that simply matching point sites with much lower resolution meshes or, indeed,
     80areas of artificially high friction
    8081than used here is an over simplification. Such simulations cannot capture the
    8182fine detail that so clearly affects inundation depth and flow speeds.
     
    8586\begin{table}
    8687\begin{center}
    87 \label{table:inundationAreas}
    88 \caption{$\rho_{in}$ and $\rho_{out}$ of the reference simulation and all sensitivity studies.}
    8988\begin{tabular}{|l|c|c|}
    9089\hline
     
    10099\end{tabular}
    101100\end{center}
     101\caption{$\rho_{in}$ and $\rho_{out}$ of the reference simulation and all sensitivity studies.}
     102
     103\label{table:inundationAreas}
    102104\end{table}
Note: See TracChangeset for help on using the changeset viewer.