Abstract: Globally,all official disaster management systems rarely focus in possible destructiveevents of tsunami risks because of the lack of information in historical data.Frequent hazards like floods and cyclones are considered in assessing naturalhazards. Return period of huge tsunami’s are not predictable, it may bethousands of years but in the recent past there were two disastrous tsunamisoccurred in 2004 and 2011 in Indian ocean and west pacific ocean respectively.
Totallyboth caused more than 250,000 fatalities and about LKR 40 trillion direct monetarylosses. Under water earth quakes are the main cause for worst tsunamioccurrences in the past. Landslides and volcanic eruption also can cause worstscenarios. So those sources also need to be considered for tsunami hazardassessment. For that numerical modelling of tsunami are needs to be done withfine inputs. By inputting the details of possible tsunami events withoutuncertainty and accurate bathymetry details will result the propagation oftsunami wave and inundation in the dry lands. From these results a hazardinnundation graph will represent the Exceedance probability for various magnitudesof tsunami thresholds. In this paper, the actual activities done by variousparties in past to find these results are described.
Especially this focuses onanalyses made for 2004 Indian Ocean tsunami in eastern coast of Sri Lanka andabout the sources needed to set up numerical computer based models with high–resolutioninputs to get the outputs of tsunami generation, propagation and inundationdetails, with which to research which sources among earth quake, landslides andvolcanic eruption may govern the inundation in the dry land area in coastalareas of Batticaloa, Ampara and Trincomalee districts.Keywords: Disaster; Earthquake; Tsunami; Landslides; Volcanic eruption; Numerical modelling; Inundation graph 1. IntroductionIn the recent past possible giant tsunamiwaves in eastern coast of Sri Lanka is mostly caused due to the Techtronicplate movement in sunda trench near Sumatra. Sunda trench is around 3200 kilometresin length and the maximum depth is around 7.7 kilometres in a particularlocation. As this is a much larger trench with various possible locations forearth quakes and landslides it was much harder to give a fine database whichconsist all the possible tsunami occurrences by different sources.
Theintention of the research is to provide a hazard curve which clearly representsthe inundation height in the dry lands of eastern coast. Previously in even itlots of improved precautions (Eg. Sea walls) were made before 2011 Japantsunami a large amount of monetary losses occurred because of this poor estimationof inundation height and the distance in the dry land. It tends for variousresearched on hazard map to explicitly indicate the uncertainty of tsunami hazardassessment. These errors were caused due to lack of good knowledge in thesource mechanisms (Non accuracy of the parameters of earthquakes and submergedlandslides), difficulties in finding the return period of these critical events,limitations of the data input and the human errors while approximating the dataon the model.
While these past procedures were done using worst scenarios, a needfor database with hazard analysis parameters which includes all possible tsunamisources and the exceedance rate is becoming essential in this world.2. Hazard AssessmentProbabilistic tsunami hazard assessment(PTHA) should assess by inputting bathymetry and topography data of the easterncoast and the actual sources of them. Numerical modelling can be done usingcommit model interface which is developed by NOAA – National Oceanic atmosphericadministration. Data can be collected from various parties as it is belongs toeither manual or satellite surveys. Gridded bathymetry data of Indian ocean canbe collected from Sri Lanka survey department or by paying for third partysources like GEBCO – General Bathymetry chart of the ocean or any othergeological survey institutions.
Description TABLE 1: Details of Input data for numerical model Input data Where to find Details Bathymetry & Topography Survey Department, GEBCO Eastern coastal area Sources Geological Institutions Along Sunda Trench Past worst case scenarios Meterological department, Marine, EIA institutions and geophysical associations At least after 1900s Effect of previous tsunmais Local Government orgganizations, Disaster management ministry and non profittable organizations worked during tsunami relief activities For 2004 Tsunami Numerical Sample Models NOAA, Engineering Community For 2004 Tsunami If data of sources are given, it can beintegrated into the model or from the PTHA observed results, the database canbe checked. FIGURE 1: The above sketch shows the mainscenarios when a worst case tsunami occurs due to the earth quake. The faultslip’s edge location is known as sunda trench. By inputting the topography andbathymetry details the model out put will show the propagation od wave and the inundationin the dry land.
3. Computer Modelling3.1. Brief DescriptionTo model the possible tsunami occurrencesand from that to get the PTHA results a software of ComMIT will be used. It wasdeveloped after the 2004 tsunami by national oceanic atmospheric administration(NOAA) in pacific marine environmental laboratory. Commonly community model isfreely available in the internet. This will provide the tools to develop modelsfor different cases in finding inundation maps in dry land area in real-timetsunami forecast applications.
The access for the ComMIT is given freelyto Indian Ocean countries. So from local databases any government organizationsor others in this region can access easily. This is giving great opportunitiesin keeping geo-spatial data locally for the secure input purposes and it allowsevery people to analyse and get the results so those results can be integratedand a reliable database can be created. Most importantly it is creating a global community of modellerswho are using same approaches so thatthey can share the information within them.3.2.
ComMIT ProcedureModels with inundation and propagation detailsshould be created using the input mentioned in TABLE 1.To access the ComMIT interface it is essentialto go through open-source software. It will lead to present the results easily.And ComMIT was written in JAVA programming language so NetCDF format will bethe input for ComMIT interface. For this commonly used NetCDF format provides severalthird party models.
For this research MOST model will be used. Followingprocedure will show the way to develop and analyse a model using numericalmethods.1. Deep water (Indian Ocean) andEastern coastal topography2. Boundary conditions in bothends – Trench and Coastal area3. Scale of model, Time-step,required resolution and the period of model run will be inputted in ComMIT.The ComMIT model interface allows theusers to share the bathymetry data and the analysed PTHA databases with otherComMIT users.
ComMIT also has a platform to display the PTHA output results inanimated format.4. ConclusionProviding the numerically analysedresults and a database with all PTHA results will incense the government to carryout the fastest way to alert the people and to develop good early warningsystems without any wrong confusion.
In future it will lead the researcher toinvent real-time monitoring systems of tsunami generation and propagation. Thiswill be a great revolution. And if geotechnical side can able to predict the timeand failure criteria for a trench with magnitude then it will be easy topredict the worst case scenarios of tsunamis. So in my words, we can shift thepeople very earlier before disaster period.
5. AcknowledgementsMany data of previous researches andsome topographical details was given by Miss. Udayanga Edirisooriya (Instructorat Civil Engineering Department) I would like to thank her for the sources.Lecturer Mr.Harsha Ratnasooriya sir guided me in collecting the references andthe details of previous literatures.
I would like to thank him too.