Geo-information for Tsunami Susceptibility Zoning -Examples from the coast of Indonesia

Michiel Ch. J. Damen
Department of Earth Systems Analysis, International Institute for Geo-information Science and Earth
Observation (ITC), Enschede, 7500 AA, the Netherlands
Phone: +31 (0) 53 4874 266 –
Web: http://www.itc.nl/news_events/tsunami.asp
Email: damen@ itc.nl

ABSTRACT
Immediately after the tsunami disaster of December 26, 2004, proposals were launched for the mitigation and even prevention of the effects of future run-up waves along the coasts of the Indian Ocean. For this, up to date tsunami susceptibility and inundation maps will be needed.

The paper describes the methodology and first results of a stepwise assessment of the coastline for tsunami susceptibility at different scales, zooming in from a national until a local level. For the national scale, Verstappen’s Geomorphological map of Indonesia is compared with the classified results of GlobeDEM elevation data, resulting in four tsunami susceptibility classes. For the regional analysis classified SRTM elevation data are merged with Landsat ETM imagery. Also here four tsunami susceptibility classes have been used for the zoning of the coastline. The detailed analysis is worked out from the Lhong area, southwest of Banda Aceh, using Quickbird and classified SRTM elevation data. The tsunami inundation zone, as delineated from the post tsunami Quickbird image of 2 January 2005, matched quite well with the 0 – 20 elevation zone.

All analysis is done in a GIS environment using freely available tsunami data from NOAA, GlobeDEM and SRTM elevation data, in combination with Landsat ETM satellite imagery.

1. INTRODUCTION
The earthquake that occurred on 26 December 2004 West of Sumatra caused a devastating tsunami, resulting in a high number of casualties in many countries around the Indian Ocean. This was especially the case in the densely populated coastal cities close to the epicenter in the Indonesian Province of Aceh, such as Banda Aceh and Meulaboh. In the city of Banda Aceh for instance, tsunami run-up waves of locally over 17 meter high reached the coast only 31 minutes after the quake (Hamzah Latief, 2004). Immediately after the tsunami disaster, proposals were launched for the mitigation of the effects of future run-up waves along the low coasts, such as the installation of a tsunami warning system (ICG/IOTWS (2005), together with “tsunami-save” reconstruction planning of the damaged areas by Provincial and National planning organizations.

Geo-information can provide the essential data for the sustainable planning for tsunami disaster risk reduction. An important mapping component in the disaster planning will be tsunami susceptibility zoning. Fortunately, immediately after 26 December 2004 a large amount of data has become available from different sources, such as satellite imagery before and after the event from RMS-SERTIT, CRISP, and others. In particular the very high resolution IKONOS and Quickbird images (Pacific Disaster Center) were very useful for a detailed analysis of the disaster zone and the Landsat ETM and ASTER data, some which have been provided for free.

The objective of the research is to develop a methodology a stepwise assessment of the coastline for tsunami susceptibility in Indonesia, with emphasis on coastal (geo)morphology and the use of medium to high resolution satellite images and digital elevation models.

2. TERMINOLOGY
Hazard Assessment is one of the components of Disaster Risk Management. By creating maps of the hazard zone, mitigation of hazardous impacts can be carried out more effectively. For instance, the evacuation planning of people after a tsunami warning should be based on a proper tsunami flood hazard zoning. According to the definition, each hazard is characterized by its location, intensity and probability (UN-ISDR, 2004). As both the intensity and probability of the past tsunami events can not be extracted in sufficient detail from the existing databases (Tsunami Data at NGDC), the term tsunami susceptibility zoning is preferred for this study, defined as a qualitative rating for the location, intensity and probability of the hazard.

3. TSUNAMI OCCURRENCE IN INDONESIA
Although the 2004 tsunami event can be considered to be the worst disaster in respect to extend, magnitude and number of casualties in history, tsunamis have also occurred frequently in the past. In Indonesia for instance, after the year 1900 more than 70 wave run-ups are know with in total over 7000 deaths, with most events in the Moluccas and Irian Jaya(Source: Tsunami Data at NGD / NOAAC) . In Figure 1 all tsunami run-ups in Indonesia after 500 A.C. are displayed, together with the epicenters of earthquakes which have triggered tsunamis, including, plate boundaries). The map shows, that tsunami events are only limited to the coast of Sumatra, but spread out all over Indonesia, except the coast of Kalimantan and Southeast Irian Jaya.

Figure 1. Tsunami run-ups and significant quakes in Indonesia after 500 A.C. (Tsunami Data at NGDC)

4. TSUNAMI SUSCEPTIBILITY ZONATION – NATIONAL SCALE
The Geomorphological map of Indonesia (Verstappen,2000) and the GlobeDEM 30” elevation model (GLOBE Project of NGDC ) have been analyzed and compared for the purpose of a “crude” tsunami susceptibility zoning of the entire coastline of Indonesia. To be able to compare surface areas, both maps have been given a Lambert Azimuthal Equal-area projection. The tsunami susceptibility rating based on the geomorphological units of Verstappen’s map is highly qualitative. Three classes have been made, based on the assumed elevation of the terrain along the coast: into Low, Medium and High (Table 1).

For the tsunami susceptibility zoning based on elevation, only four tsunami susceptibility classes are made using the GlobeDEM 1m.elevation model into Low, Moderately High, High & Very High. (Table 2) The 30” pixels have been resampled to a grid of 1x1 km. By comparing the classified results of both maps in Figure 2 a & b, it can be concluded that the Unit Alluvial Plains in the Geomorphological have in general a High to Very High susceptibility class in the GlobeDEM

Figure 2 a Geomorphological Map – Aceh-
Scale approx.. 1: 5.000.000 (Verstappen, 2000)

Figure 2 b GlobeDEM Aceh. - Scale approx..1:5.000.000 - See also Table 2.

Table 1.

Table 2 GlobeDEM classification - National scale

map, and that geomorphological units as Hills, Mountains and Volcanic cones are classified as Moderately High based on elevation only. However, when comparing the geographic position of the classified coastline stretches, large deviations can be seen. This may be caused by the relative small original scale (1:5.000.000) of the Geomorphological map; the GlobeDEM has a higher positional accuracy.

It is therefore recommended to use as a first step in the susceptibility classification the GlobeDEM data, and in a second step Verstappen geomorphological map for a better understanding of the landforms and processes in the different elevation classes.

5. TSUNAMI SUSCEPTIBILITY ZONATION – REGIONAL SCALE
After a crude estimation of the tsunami susceptibility classes on national scale, a more detailed analysis can be done with higher resolution data, such as Landsat Enhanced Thematic Mapper satellite imagery in combination with Shuttle Radar Topography Mission (SRTM) elevation data (SRTM, NASA). The classification methodology is similar to those of the national scale. In Figure 3 the Landsat ETM image is “draped” in a transparent way over the classified SRTM elevation model.

Table 3. SRTM & Landsat ETM Regional scale classification

Figure 3. Merge of classified SRTM with Landsat ETM of 15 August 2001

6. TSUNAMI SUSCEPTIBILITY ZONATION – LOCAL SCALE
Using SRTM elevation data in combination with very high resolution satellite imagery, such as IKONOS or Quickbird data, a tsunami susceptibility zoning on local scale can be done. In Figure 4 the classified SRTM elevation data have been transparently “draped” over 2 January 2005 (post tsunami) Quickbird image of the Lhong area, located southwest of the city of BandaAceh. In the classification on regional scale, this area was classified as highly susceptible.

The red line indicates the maximum extend of the inundation and the blue line gives the pre-tsunami coastline, digitized from a topographical map, scale 1:50.000. By comparing the 10 - 20 meter elevation zone with the maximum inundation, it can be concluded that they fit quite well in this type of coastal morphology.

Figure 4 Inundation zone – Lhong, Aceh Drape of Quickbird of 02 January 05 over classified SRTM elevation data

7. CONCLUSIONS & RECOMMENDATIONS

• For an assessment of tsunami susceptibility on national scale, classified GlobeDEM elevation data with 1 km 2 pixels can give a first “crude” result. For a better knowledge of the landforms of the different susceptibility classes the use of a geomorphological map is recommended.
• For tsunami hazard zoning more detailed data on the frequency of occurrence will be needed.
• The merge of SRTM elevation data (~92 m cells) with satellite images like Landsat ETM+ or ASTER is suitable for a tsunami susceptibility classification on regional scale. It is recommended to include in the analysis not only the elevation, but also the presence of (fringing) coral reefs, mangrove forest and coastline protection works.
• After the merge of classified SRTM elevation data with the post tsunami Quickbird image of the Lhong area in Aceh, it can be concluded that by the 20 meter elevation contour fits quite well with the maximum extend of the inundation.
• For more detailed tsunami susceptibility studies the NASA should provide its 30 m. SRTM data.

8. DATA TYPES & SOURCES

9. REFERENCES

• GLCF – Global Land Cover Facility - http://www.modis-cluster.org/data/
• GLOBE Project of NGDC - NOAA Satellite and Information Service - http://www.ngdc.noaa.gov/mgg/topo/globe.html
• CRISP – Center for Remote Sensing and Processing – Univ. of Singapore - http://www.crisp.nus.edu.sg/tsunami/tsunami.html
• Hamzah Latief (2004) – Tsunami Aceh 2004 – PowerPoint presentation - Tsunami Research Group, Institut Teknologi, Bandung, Indonesia
• ICG/IOTWS (2005) - Towards the Establishment of a Tsunami Warning and Mitigation System for the Indian Ocean http://ioc.unesco.org/indotsunami/index.htm
• Pacific Disaster Center - http://www.pdc.org/apnhintsu/html/apnhintsu-init.jsp
• RMS-SERTIT - Rapid Mapping Service – Sertit - http://sertit.u-strasbg.fr/documents/asie/indonesia.htm
• SRTM Data NASA - ftp://e0mss21u.ecs.nasa.gov/srtm/
• Tsunami Data at NGDC - NOAA Satellite and Information Service - http://www.ngdc.noaa.gov/seg/hazard/tsu.shtml
• UN- ISDR (2004) UN International Strategy for Disaster Reduction - Terminology: Basic terms of disaster risk reduction - http://www.unisdr.org/eng/library/lib-terminology-eng%20home.htm
• Verstappen, H. Th. (2000) - Outline of the Geomorphology of Indonesia – ITC, Enschede, the Netherlands.