Changeset 7522


Ignore:
Timestamp:
Sep 22, 2009, 6:09:32 PM (15 years ago)
Author:
ole
Message:

Incorporated review comments from Jane. Waiting on new Jason figure and reply from DB.

Location:
anuga_work/publications/boxing_day_validation_2008
Files:
9 edited

Legend:

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

    r7480 r7522  
    1616  provided by the \textsc{ursga} model. The left image shows the maximum
    1717  modelled depth while the right hand image shows the maximum modelled
    18   flow velocities.}
     18  flow speeds.}
    1919\label{fig:reference_model}
    2020\end{figure}
     
    3434  friction value while the higher slows the flow and decreases the
    3535  inundation extent. Ideally, friction should vary across the entire
    36   domain depending on terrain and vegetation, but this is beyond the
     36  domain depending on terrain and vegetation. This, however, is beyond the
    3737  scope of this study.}
    3838\label{fig:sensitivity_friction}
     
    4444\end{center}
    4545
    46 \caption{The maximal flow speeds for the same model parameterisations
     46\caption{The maximal modelled flow speeds for the same model parameterisations
    4747  found in Figure \protect \ref{fig:sensitivity_friction}.
    4848  The reference flow speeds for a
  • anuga_work/publications/boxing_day_validation_2008/conclusion.tex

    r7521 r7522  
    11\section{Conclusion}
    22This paper proposes a new field data benchmark for the
    3 verification of tsunami inundation models. Currently, there is a
     3validation of tsunami inundation models. Currently, there is a
    44scarcity of appropriate validation datasets due to a lack of well-documented
    55historical tsunami impacts. The benchmark proposed here
     
    2020and able to predict detailed inundation extents and dynamics with reasonable accuracy.
    2121Model predictions matched well a detailed inundation survey
    22 of Patong Bay, Thailand as well as altimetry data from the \textsc{jason} satellite,
     22of Patong City, Thailand as well as altimetry data from the \textsc{jason} satellite,
    2323eye-witness accounts of wave front arrival times and onshore flow speeds.
    2424
  • anuga_work/publications/boxing_day_validation_2008/data.tex

    r7521 r7522  
    77subsequent coastal field surveys of run-up and flooding, and
    88measurements of coseismic displacements as well as bathymetry from ship-based
    9 expeditions, have now been made
     9expeditions and high quality topographic data, have now been made
    1010available. %~\cite{vigny05,amnon05,kawata05,liu05}.
    1111
     
    2222cause of any discrepancies between modelled and observed inundation.
    2323Consequently, in this section we present data not only to facilitate
    24 validation of inundation but to also aid the assessment of tsunami
     24validation of inundation extent but to also aid the assessment of tsunami
    2525generation and propagation.
    2626
     
    6969it encounters the shoreline bordering coastal regions. This period
    7070of the tsunami evolution is referred to as the propagation stage. The
    71 height and velocity of the tsunami is dependent on the local
     71height and speed of the tsunami is dependent on the local
    7272bathymetry in the regions through which the wave travels and the size
    7373of the initial wave. This section details the bathymetry data needed
     
    7979sources:
    8080\begin{itemize}
    81 \item a two arc minute grid data set covering the Bay of Bengal,
     81\item a two arc minute data grid covering the Bay of Bengal,
    8282  DBDB2, obtained from US Naval Research Labs
    8383  (\url{http://www7320.nrlssc.navy.mil/DBDB2_WWW});
    84 \item a 3 second arc grid obtained directly from NOAA covering the
     84\item a three arc second data grid obtained directly from NOAA covering the
    8585  whole of the Andaman Sea based on the
    86   Smith \& Sandwell 2-minute
     86  Smith \& Sandwell two minute
    8787  dataset (\url{http://topex.ucsd.edu/WWW_html/srtm30_plus.html}),
    8888  coastline constrained using SRTM data (\url{http://srtm.csi.cgiar.org})
    8989  as well as Thai Navy charts no.\ 45 and no.\ 362; and 
    90 \item a one second grid created from the digitised Thai Navy
    91   bathymetry chart, no. 358, which covers Patong Bay and the
    92   immediately adjacent regions. The digitised points and contour lines
    93 from this chart are shown in Figure~\ref{fig:patong_bathymetry}.
    94   The gridding of data was performed using \textsc{Intrepid}, a commercial
    95   geophysical processing package developed by Intrepid Geophysics. The
    96   gridding scheme employed the nearest neighbour algorithm followed by
    97   an application of minimum curvature akima spline smoothing.
    98   See \url{http://www.intrepid-geophysics.com/ig/manuals/english/gridding.pdf}
    99   for details on the Intrepid gridding scheme. 
     90\item Thai Navy chart no.\ 358 providing water depths in Patong Bay.
    10091\end{itemize}
    10192
    102 These sets were combined via
    103 interpolation and resampling to produce four nested grids
    104 which are relatively coarse in the deeper water and
    105 progressively finer as the distance to
    106 Patong Beach decreases as shown in Figure~\ref{fig:nested_grids}. 
    107 
    108 The coarsest
    109 bathymetry was obtained by interpolating the DBDB2 grid to a 27 second
    110 arc grid. A subsection of this region was then replaced by nine second
    111 data which was generated by sub-sampling the three second of arc grid from
    112 NOAA. It is an artificially generated data set which is a subset of the original data.
    113 
    114 A subset of the nine second grid was replaced by the three second
    115 data. Finally, the one second grid was used to approximate the
    116 bathymetry in Patong Bay. Any
    117 points that deviated from the general trend near the boundary were
    118 deleted as a quality check.
    119 
    120 A one second grid was used to approximate the bathymetry in Patong
    121 Bay. This bathymetry data was created from the digitised Thai
    122 Navy bathymetry chart, no 358.
    123 
    124 
     93These data sets were combined via gridding, interpolation and resampling to produce
     94four nested grids which are relatively coarse in the deeper water and
     95progressively finer as the distance to shore Patong Beach decreases as
     96shown in Figure~\ref{fig:nested_grids}. This progression was chosen
     97to match model resolution requirements according to the principle that
     98shallow water flows are more sensitive to variations in elevation data
     99than deep water flows. Consequently, the elevation data in shallow
     100waters and on-shore need to be resolved better than elevation data
     101further off-shore.
     102 
     103The coarsest bathymetry was obtained by interpolating the DBDB2 grid
     104to a 27~second arc grid. A subsection of this region was then replaced
     105by nine second data which was generated by sub-sampling the three
     106second of arc grid from NOAA. It is an artificially generated data set
     107which is a subset of the original data.  A subset of the nine second
     108grid was replaced by the three second data. Finally, a one arc second
     109grid approximating the bathymetry in Patong Bay and the immediately
     110adjacent regions was created by digitising Thai Navy bathymetry chart,
     111no.\ 358. The digitised points and contour lines from this chart are
     112shown in Figure~\ref{fig:patong_bathymetry}. The gridding was
     113performed using \textsc{Intrepid}, a commercial geophysical processing
     114package developed by Intrepid Geophysics\footnote{
     115See
     116\url{http://www.intrepid-geophysics.com/ig/manuals/english/gridding.pdf}
     117for details on the Intrepid gridding scheme.}.
     118Any points that deviated from the general trend near the boundary were
     119deleted through a quality control process.
    125120The sub-sampling of larger grids was performed by using \textsc{resample},
    126121a Generic Mapping Tools (\textsc{GMT}) program \cite{wessel98}.
     
    132127\end{center}
    133128
    134 \caption{Nested bathymetry grids.}
     129\caption{Nested elevation grids of the Andaman Sea with
     130highest resolution at and around Patong Bay.}
    135131\label{fig:nested_grids}
    136132\end{figure}
     
    138134\subsubsection{JASON Satellite Altimetry}\label{sec:data_jason}
    139135During the 26 December 2004 event, the \textsc{jason} satellite tracked from
    140 north to south and over the equator at 02:55 UTC nearly two hours
     136north to south and over the equator at 02:55~UTC nearly two hours
    141137after the earthquake \cite{gower05}. The satellite recorded the sea
    142138level anomaly compared to the average sea level from its previous five
     
    162158Coordinating Committee Co-ordinating Committee for Geoscience
    163159Programmes in East and Southeast Asia (CCOP) \cite{szczucinski06}
    164 was obtained to validate model inundation. See also acknowledgements
    165 at the end of this paper. In this section we also present eye-witness
     160was obtained to validate model inundation. In this section we also present eye-witness
    166161accounts which can be used to qualitatively validate tsunami
    167162inundation.
     
    172167(described in Section \ref{sec:bathymetry data}) and from 1~m and 10~m
    173168elevation contours provided by the CCOP. The 1~second terrain model
    174 for the and community as shown in Figure~\ref{fig:patong_bathymetry}.
    175 
    176 Two 1/3~second grids were created: One for the saddle point covering
    177 Merlin and Tri Trang Beaches and one for Patong City and its immediate
    178 shore area.  These grids were based on the same data used for
     169for the community is shown in Figure~\ref{fig:patong_bathymetry}.
     170
     171To provide increased resolution for the surveyed area,
     172two 1/3~second grids were created: One for the saddle point covering
     173Merlin and Tri Trang Beaches (separate survey patch to the left in
     174Figure~\ref{fig:patongescapemap})
     175and one for Patong City and its immediate
     176shore area (main surveyed area in Figure~\ref{fig:patongescapemap}).
     177These grids were based on the same data used for
    179178the 1~second data grid.  The Patong city grid was further modified based on
    180179satellite imagery to include the river and lakes towards the south of
     
    188187\end{center}
    189188
    190 \caption{3D visualisation of the elevation data set used for the nearshore propagation and inundation in Patong Bay showing
     189\caption{3D view of the elevation data set used for the nearshore propagation and inundation in Patong City showing
    191190digitised data points and contours as well as rivers and roads
    192191draped over the data model.}
     
    197196\subsubsection{Buildings and Other Structures}
    198197Human-made buildings and structures can significantly affect tsunami
    199 inundation. The footprint and number of floors of the
    200 buildings in Patong Bay were extracted from the data provided by CCOP.
    201 The heights of these
    202 buildings were estimated assuming that each floor has a height of 3 m and they
    203 were added to the topographic dataset.
     198inundation. The footprint and number of floors of the buildings in
     199Patong Bay were extracted from the data provided by CCOP.  The heights
     200of these buildings were estimated assuming that each floor has a
     201height of 3~m and the resulting profiles were added to the topographic
     202dataset. The resulting elevation model and its interaction with one of
     203the tsunami waves can be seen in Figure~\ref{fig:anuga screenshot} in
     204Section~\ref{sec:inundation results}.
    204205
    205206
     
    225226
    226227\caption{Tsunami survey mapping the maximum observed inundation at
    227   Patong beach courtesy of the CCOP \protect \cite{szczucinski06}.}
     228  Patong City courtesy of the CCOP \protect \cite{szczucinski06}.}
    228229\label{fig:patongescapemap}
    229230\end{figure}
     
    246247\end{center}
    247248
    248 \caption{Location of timeseries extracted from the model output.}
     249\caption{Location of time series extracted from the model output.}
    249250\label{fig:gauge_locations}
    250251\end{figure}
     
    255256(Comfort Resort) and
    256257\url{http://www.archive.org/download/tsunami_patong_beach/tsunami_patong_beach.wmv}
    257 (Novotel)}
     258(Novotel).}
    258259%http://wizbangblog.com/content/2005/01/01/wizbang-tsunami.php
    259 which include footage of the tsunami in Patong Bay on the day
    260 of the 2004 Indian Ocean Tsunami. Both videos show an already inundated
     260which include footage of the tsunami in Patong City on the day
     261of the 2004 Indian Ocean tsunami. Both videos show an already inundated
    261262street. They also show what is to be assumed as the second
    262263and third waves approaching and further flooding of the buildings and
    263 street.  The first video is in the very north, filmed from what is
     264street. The first video is in the very north, filmed from what is
    264265believed to be the roof of the Novotel Hotel marked ``north'' in Figure
    265266\ref{fig:gauge_locations}. The second video is in the very south,
     
    304305should reproduce the following behaviour:
    305306\begin{itemize}
    306  \item reproduce the inundation survey map in Patong city
     307 \item the inundation survey map in Patong city
    307308   (Figure~\ref{fig:patongescapemap}),
    308  \item simulate a leading depression followed by two distinct crests
    309    of decreasing magnitude at the beach, and
     309 \item a leading depression followed by two distinct crests
     310   of decreasing magnitude at the beach,
    310311 \item predict the water depths and flow speeds, at the locations of
    311312   the eye-witness videos, that fall within the bounds obtained from
    312    the videos.
    313  \item reproduce the \textsc{jason} satellite altimetry sea surface
    314    anomalies (see Section~\ref{sec:data_jason}),
    315  \item reproduce the vertical deformation observed in north-western
     313   the videos,
     314 \item the \textsc{jason} satellite altimetry sea surface
     315   anomalies (see Section~\ref{sec:data_jason}), and
     316 \item the vertical deformation observed in north-western
    316317   Sumatra and along the Nicobar--Andaman islands (see
    317    Section~\ref{sec:gen_data}),
     318   Section~\ref{sec:gen_data}).
    318319\end{itemize}
    319320
    320 Ideally, the model should also be compared to measured timeseries of
    321 waveheights and velocities but the authors are not aware of the
     321Ideally, the model should also be compared to measured time series of
     322wave heights and flow speeds but the authors are not aware of the
    322323availability of such data near Patong Bay.
  • anuga_work/publications/boxing_day_validation_2008/introduction.tex

    r7521 r7522  
    2020These models are typically used to predict quantities such as arrival
    2121times, wave speeds and heights, as well as inundation extents
    22 which can be used to develop efficient hazard mitigation plans. Physics based
    23 models combine observed seismic, geodetic and sometimes tsunami data to
     22that can be used to develop efficient hazard mitigation plans. Physics based
     23models combine observed seismic, geodetic and sometimes tide gauge data to
    2424 provide estimates of initial sea floor and ocean surface
    2525deformation. The shallow water wave equations~\cite{george06},
     
    2929
    3030Inaccuracies in model prediction can result in inappropriate
    31 evacuation plans and town zoning, which may result in loss of life and
    32 large financial losses. Consequently tsunami models must undergo
     31evacuation plans, town zoning and land use planning,
     32which ultimately may result in loss of life and infrastructure.
     33Consequently tsunami models must undergo
    3334sufficient end-to-end testing to increase scientific and community
    3435confidence in the model predictions.
     
    6364
    6465Currently, the extent of tsunami-related field data is limited. The
    65 cost of tsunami monitoring programs as well as
     66cost of tsunami monitoring programs and the rarity of events as well as
    6667bathymetry and topography surveys
    6768prohibits the collection of data in many of the regions in which
     
    9697al~\cite{synolakis08} to validate and verify tsunami models.
    9798The benchmark proposed here allows evaluation of
    98 model structure during all three distinct stages tsunami evolution.
     99model components during three distinct stages tsunami
     100evolution, namely generation, propagation and inundation.
    99101It consists of geodetic measurements of the
    100102Sumatra--Andaman earthquake that are used to validate the description
     
    115117 tailored accordingly.
    116118
    117 Unlike the existing field benchmarks the proposed test facilitates
     119Unlike the existing field benchmarks, the proposed test facilitates
    118120 localised and highly detailed spatially distributed assessment of
    119121modelled inundation. To the authors knowledge it is also the first benchmark to
     
    130132The numerical models used to simulate tsunami impact
    131133are computationally intensive and high resolution models of the entire
    132 evolution process will often take a number of days to
    133 run. Consequently, the uncertainty in model predictions is difficult to
    134 quantify as it would require a very large number of runs.
     134evolution process will often require a number of days to
     135complete. Consequently, the uncertainty in model predictions is difficult to
     136quantify as it would require a very large number of simulations.
    135137However, model uncertainty should not be ignored. Section
    136138~\ref{sec:sensitivity} provides a simple analysis that can
    137139be used to investigate the sensitivity of model predictions to a number
    138 of model parameters.
     140of key model parameters.
  • anuga_work/publications/boxing_day_validation_2008/method.tex

    r7521 r7522  
    11\section{Modelling the Event}\label{sec:models}
    22Numerous models are currently used to model and predict tsunami
    3 generation, propagation and run-up. These range in solving different
     3generation, propagation and inundation. These range in solving different
    44equations and employing different methodologies with some examples
    55being~\cite{titov97a,satake95,zhang08}. Here we introduce the
     
    6565~\cite{chlieh07,asavanant08,arcas06,grilli07,ioualalen07}. Some are
    6666determined from various geological surveys of the site. Others solve
    67 an inverse problem which calibrates the source based upon the tsunami
     67an inverse problem that calibrates the source based upon the tsunami
    6868wave signal, the seismic signal and/or even the run-up.
    6969The source
     
    7373range of geodetic and seismic data. The slip model consists
    7474of 686~20~km~x~20~km subsegments each with a different slip, strike and dip
    75 angle. The dip subfaults go from $17.5^\circ$ in the north and $12^\circ$ in
     75angle. The dip subfaults range from $17.5^\circ$ in the north and $12^\circ$ in
    7676the south. Refer to Chlieh et al~\cite{chlieh07} for a detailed
    7777discussion of this model and its derivation. %Note that the geodetic
     
    8484%accurately.
    8585
    86 \subsection{Deep water propagation}\label{sec:modelPropagation}
     86\subsection{Open water propagation}\label{sec:modelPropagation}
    8787The \textsc{ursga} model described below was used to simulate the
    8888propagation of the 2004 Indian Ocean tsunami across the open ocean, based on a
     
    107107grid system described in Section \ref{sec:propagation data} where
    108108the coarse grids were used in the open ocean and the finest
    109 resolution grid was employed in the region closest to Patong bay.
     109resolution grid was employed in the region closest to Patong City.
    110110\textsc{Ursga} is not publicly available.
    111111
     
    115115unless an intricate sequence of nested grids is employed. In
    116116comparison \textsc{anuga}, described below, is designed to produce
    117 robust and accurate predictions of on-shore inundation, but is less
     117robust and accurate predictions of inundation, but is less
    118118suitable for earthquake source modelling and large study areas because
    119119it is based on projected spatial coordinates. Consequently, the
  • anuga_work/publications/boxing_day_validation_2008/paper.tex

    r7521 r7522  
    1414%----------title-------------%
    1515\title{Benchmarking Tsunami Models using the December 2004 Indian
    16   Ocean Tsunami and its Impact at Patong Bay}
     16  Ocean Tsunami and its Impact at Patong City, Thailand}
    1717\titlerunning{A tsunami model benchmark}
    1818
     
    2828        \email{john.jakeman@anu.edu.au}
    2929        \and
    30         O. Nielsen \and R. Mleczko \and D. Burbidge \and K. Van Putten \and N. Horspool \at
     30        O. Nielsen \and K. Van Putten \and R. Mleczko
     31        \and D. Burbidge \and N. Horspool \at
    3132        Geoscience Australia, Canberra, \textsc{Australia}
    3233}
     
    5354propagation and a detailed inundation survey of Patong city, Thailand
    5455to compare model and observed inundation. Furthermore we utilise this
    55 benchmark to further validate the \textsc{ursga--anuga} modelling methodology
     56benchmark to further validate the modelling methodology
    5657 used by Geoscience Australia to simulate the tsunami
    5758inundation. Important buildings and other structures were incorporated
    58 into the underlying computational mesh and shown to have a large
     59into the underlying computational mesh and are shown to have a large
    5960influence on inundation extent.
    6061
  • anuga_work/publications/boxing_day_validation_2008/results.tex

    r7521 r7522  
    118118After propagating the tsunami in the open ocean using \textsc{ursga},
    119119the approximated ocean and surface elevation and horizontal flow
    120 velocities were extracted and used to construct a boundary condition
     120speeds were extracted and used to construct a boundary condition
    121121for the \textsc{anuga} model. The interface between the \textsc{ursga}
    122122and \textsc{anuga} models was chosen to roughly follow the 100~m depth
    123 contour along the west coast of Phuket Island. Data from the 
    124 3 second grid was decimated to match the resolution chosen in ANUGA.
    125 The computational
     123contour along the west coast of Phuket Island. Data from the
     124three~second grid which is approximately 30~m apart was decimated to
     125match the resolution chosen in \textsc{Anuga}.  The computational
    126126domain is shown in Figure~\ref{fig:computational_domain}.
    127127
     
    164164The boundary condition at each side of the domain towards the south
    165165and the north where no information about the incident wave or
    166 its velocity field is available
     166its velocity field is available from the \textsc{Ursga} model
    167167was chosen as a transmissive
    168168boundary condition, effectively replicating the time dependent wave
     
    170170The velocity field on these boundaries was kept at
    171171to zero during the simulation. Other choices include applying the mean tide value as a
    172 Dirichlet boundary condition. But experiments as well as the
     172Dirichlet boundary condition. Experiments as well as the
    173173result of the verification reported here showed that this approach
    174174tends to underestimate the tsunami impact due to the tempering of the
     
    176176condition robustly preserves the wave.
    177177
    178 During the \textsc{anuga} simulation the tide was kept constant at
     178During the \textsc{anuga} simulation the tide was kept constant in the offshore region at
    179179$0.80$ m. This value was chosen to correspond to the tidal height
    180180specified by the Thai Navy tide charts
     
    184184wave propagated through the \textsc{anuga} domain is much
    185185smaller of the order of 2 hours. Consequently the assumption of constant tide height is
    186 reasonable.
     186reasonable. The initial water level for the river was set to 0.
    187187
    188188\subsection{Inundation}\label{sec:inundation results}
    189189The \textsc{anuga} simulation described in the previous section and used to
    190190 model shallow water propagation also predicts
    191 inundation. Maximum onshore inundation depth was computed from the model
    192 throughout the entire Patong Bay region and used to generate
    193 a measure of the inundated area.
     191inundation. Maximum onshore inundation depth was computed from the inundation model
     192and used to generate a measure of the inundated area.
    194193Figure~\ref{fig:inundationcomparison1cm} (left) shows very good
    195194agreement between the measured and simulated inundation. However,
     
    252251\end{equation}
    253252These values for the two aforementioned simulations are given in
    254 Table~\ref{table:inundationAreas}. High value of both $\rho_{in}$ and $\rho_{out}$ indicate
     253Table~\ref{table:inundationAreas} along with results from the sensitivity analysis in
     254Section~\ref{sec:sensitivity}. High values of both $\rho_{in}$ and $\rho_{out}$ indicate
    255255that the model overestimates the impact whereas low values of both quantities would indicate
    256256an underestimation. A high value of $\rho_{in}$ combined with a low value of $\rho_{out}$
     
    267267flow, as well as uncertainties in the elevation data including effects of
    268268erosion and deposition by the tsunami event.
    269 The impacts of some of the model uncertainties are is investigated in
     269The impacts of some of the model uncertainties are as investigated in
    270270Section~\ref{sec:sensitivity}.
    271271
     
    274274As one aim of this paper is to provide a new benchmark for tsunami
    275275inundation modelling we have made the datasets available
    276 available on \textsc{Sourceforge} in \textsc{anuga}
     276available on \textsc{Sourceforge} in the \textsc{anuga}
    277277project (\url{http://sourceforge.net/projects/anuga}) under the directory
    278278\url{validation\_data/patong-1.0}.
     
    284284will need to run the validation scripts (\url{anuga\_validation/automated\_validation\_tests/patong\_beach\_validation}) which are part of the
    285285\textsc{anuga} distribution also available from
    286 Sourceforge \url{http://sourceforge.net/projects/anuga}.
     286\textsc{Sourceforge} \url{http://sourceforge.net/projects/anuga}.
    287287
    288288
     
    291291\subsubsection{Arrival time}
    292292The arrival time of the first wave took place between 9:55 and 10:05 as described in
    293 Section~\ref{sec:eyewitness data}. The modelled arrival time at the beach is 10:02
     293Section~\ref{sec:eyewitness data}. The modelled arrival time at the beach is around 10:02
    294294as can be verified from the animation provided in
    295 Section \ref{sec:inundation results}.
     295Section \ref{sec:inundation results} or from Figure~\ref{fig:onshore_timeseries} below.
    296296Subsequent waves of variable magnitude appear over the next two hours
    297297approximately 20-30 minutes apart.
     
    304304series have been extracted from the model. These are the locations where video footage from the event is
    305305available as described in Section \ref{sec:eyewitness data}.
    306 The corresponding are shown in Figure \ref{fig:onshore_timeseries}.
     306The corresponding time series are shown in Figure \ref{fig:onshore_timeseries}.
    307307
    308308
     
    325325
    326326\caption{Time series obtained from the two onshore locations, North and South,
    327 shown in Figure \protect \ref{fig:gauge_locations}.}
     327shown in Figure \protect \ref{fig:gauge_locations}. Time is given in hours since the earthquake event (7:59am).}
    328328\label{fig:onshore_timeseries}
    329329\end{figure}
     
    335335Table \ref{tab:depth and flow comparisons}.
    336336The predicted maximum depths and speeds are all of the same order
    337 of what was observed. However, unlike the real event,
     337of what was observed as is the approximate arrival time at the two locations. However, unlike the real event,
    338338the model estimates complete withdrawal of the water between waves at the
    339339chosen locations and shows that the model must be used with caution at this
    340340level of detail.
    341 Nonetheless, this comparison serves to check that depths and speeds
     341Nonetheless, this comparison serves to check that the peak depths and speeds
    342342predicted are within the range of what is expected.
    343343
  • anuga_work/publications/boxing_day_validation_2008/sensitivity.tex

    r7480 r7522  
    55are computationally intensive and high resolution models of the entire
    66evolution process will often take a number of days to
    7 run. Consequently, the uncertainty in model predictions is difficult to
     7compute. Consequently, the uncertainty in model predictions is difficult to
    88quantify as it would require a very large number of runs.
    99However, model uncertainty should not be ignored. The aim of this section is
     
    2828appropriate values of Manning's coefficient range from 0.007 to 0.03
    2929for tsunami propagation over a sandy sea floor and the reference model
    30 uses a value of 0.01.  To investigate sensitivity to this parameter,
     30uses a value of 0.01. To investigate sensitivity to this parameter,
    3131we simulated the maximum onshore inundation using a Manning's
    3232coefficient of 0.0003 and 0.03. The resulting inundation maps are
     
    3434and the maximum flow speeds in Figure~\ref{fig:sensitivity_friction_speed}.
    3535The figure, along with Table~\ref{table:inundationAreas},
    36 shows that the on-shore inundation extent decreases with increasing
     36shows, as expected, that the on-shore inundation extent decreases with increasing
    3737friction and that small perturbations in the friction cause bounded
    3838changes in the output. This is consistent with the conclusions of
     
    4646generally well predicted by the generation and propagation models such as
    4747\textsc{ursga} as demonstrated in Section \ref{sec:resultsPropagation}
    48 and also in \cite{thomas2009}. Nevertheless, the effect of errors in
     48and also in \cite{thomas2009} assuming that the source parameters chosen appropriately.
     49Nevertheless, the effect of errors in
    4950the wave height used as input to the inundation model \textsc{anuga}
    5051was investigated by perturbing the
    51 amplitude of the input wave by $\pm$10 cm. This value was chosen to be larger
    52 than the expected error in the amplitude predicted by the propagation model.
     52amplitude of the input wave by $\pm$10 cm. This value was chosen to be consistent
     53with the expected error in the amplitude predicted by the propagation model
     54and amounts to about $\pm$5\% of the maximal waveheight at the boundary.
    5355
    54 Figure~\ref{fig:sensitivity_boundary}, Figure~\ref{fig:sensitivity_boundary_speed},
    55 and  Table~\ref{table:inundationAreas}
     56Figure~\ref{fig:sensitivity_boundary}, Figure~\ref{fig:sensitivity_boundary_speed}
     57and Table~\ref{table:inundationAreas}
    5658indicate that the inundation severity is directly proportional to the
    5759boundary waveheight but small
  • anuga_work/publications/boxing_day_validation_2008/tsunami07.bib

    r7521 r7522  
    10481048author = {Burbidge, D. and Cummins, P.R. and Mleczko, R. and Thio, H.K.},
    10491049title = "{A Probabilistic Tsunami Hazard Assessment for {W}estern {A}ustralia}",
    1050 journal = {Pure appl. geophys.},
     1050journal = "{Pure Appl. Geophys.}",
    10511051year = {2008},
    10521052volume = {165},
Note: See TracChangeset for help on using the changeset viewer.