In this report, impact modelling refers to casualties and
damage to residential buildings as a result
of the inundation described in Section \ref{sec:results}. It is assumed
that the event occurs at night. 
Exposure data are sourced from the National Building Exposure Database (NBED), 
developed by GA\footnote{http://www.ga.gov.au/urban/projects/ramp/NBED.jsp}. 
It contains information about residential buildings, people and the
cost of replacing buildings and contents.

To develop building damage and casuality estimates, 
residential collapse vulnerability models and casualty models were developed.
The vulnerability models have been developed for 
framed residential construction using data from the Indian Ocean tsunami event. The models predict the collapse 
probability for an exposed population and incorporates the following 
parameters known to influence building damage \cite{papathoma:vulnerability},

\begin{itemize}
\item	inundation depth at building	
\item	distance from the coast	
\item	building material (residential framed construction)	
\item	inundation depth in house above floor level
\end{itemize}	

The collapse vulnerability models used are presented in Table \ref{table:collapse}. 
%In applying the model, all structures in the inundation zone were 
%spatially located and the local water depth and building row 
%number from the exposed edge of the suburb were determined for each %structure.

Casualty models were based on the
storm surge models used for the Cairns Cyclone Scenario and 
through consultation with Dr David Cooper of NSW Health, \cite{cooper:2005}. 
The injury probabilities for exposed populations were determined 
based on the nocturnal nature of the event, the collapse outcome 
for the structure, the water depth with respect to 
sleeping height (1.0 m) and the limited warning noise for people 
in the first three city blocks (six house rows) that could potentially 
awaken them. The three injury categories correspond with the 
categories presented in HAZUS-MH \cite{NIBS:2003} for earthquake 
related injury. The casualty model used is presented in Table
\ref{table:casualty} 
and the injury categories are presented in Table \ref{table:injury}. 
Input data comprised of resident population data at census
district level derived from the ABS 2001 Census.

South Hedland is not exposed to inundation in this
scenario, we therefore restrict the damage modelling to a smaller section of the NBED.
For the damage modelling, there
are an estimated 3700 residential structures and a population of approximately
11500\footnote{Population is determined by census data and the 1999
ABS housing survey}. 
The damage to the residential structures in this section of the Port Hedland community
is summarised in Table \ref{table:damageoutput}. The percentage
of repair cost to structural value shown is based on the total structural value
of \$M. Likewise, the percentage of contents loss shown is
based on the total contents value of \$M for
the region. 
%The injuries sustained is summarised in Table \ref{table:injuries}. 
The HAT scenario is the only scenario to cause damage 
to Port Hedland with around \% of the population affected. 

\begin{table}[h]
\begin{center}
\caption{Residential damage sustained for the MSL, HAT and LAT scenarios.}
\label{table:damageoutput}
\begin{tabular}{|l|l|l|l|l|l|l|}\hline
&Houses  & Houses  & Structural & Repair Cost \% & Contents & Contents Loss \% \\ 
&Inundated & Collapsed & Repair Cost 
& of Total Value & Losses & of Total Value \\ \hline
%MSL & & 1 & \$ &   \% & \$ &  \% \\ \hline
HAT & &\$M & &\$M & & \\ \hline
%LAT & & & & & & \\ \hline
\end{tabular}
\end{center}
\end{table}

%\begin{table}[h]
%\begin{center}
%\caption{Injuries sustained for the MSL, HAT and LAT scenarios.}
%\label{table:injuries}
%\begin{tabular}{|l|l|l|l|l|l|}\hline
%&Minor & Moderate & Serious & Fatal \\ \hline
%MSL & &  &  & \\ \hline
%HAT & & & & \\ \hline
%LAT & & & & \\ \hline
%\end{tabular}
%\end{center}
%\end{table}

Tsunami impact on indigeneous communities should be considered
especially as a number of communities exist in coastal regions of north west WA.
These communities are typically not included in national residential databases 
and would be therefore overlooked in damage model estimates.

There are four indigeneous communities located in this study areal Tjalkli Warra, Jinparinya,
Punju Ngarugundi Njamal and Tjalka Boorda.
Tjalka Boorda is located in a potentially vulnerable 
position
(on the headland) whose population is not registered
\footnote{get a reference from Anita}. 
%The community is not affected for any of the scenarios (see Figure 
%\ref{fig:gaugeTjalkaBoordaAboriginalReserve}).

%\begin{center}
%\begin{tabular}{|l|l|l|l|}\hline
%Easting	 & Northing	& Community	& Population \\ \hline
%677055.85&	7742819.31&	 Tjalkli Warra&	100 \\ \hline
%690756.92&	7746148.99&	 Jinparinya&	30 \\ \hline
%691091.39&	7747119.61&	 Punju Ngarugundi Njamal&	22 \\ \hline
%669526.15&    7752820.51& Tjalka Boorda	& 0 \\ \hline
%\end{tabular}
%\end{center}