THE ASSESSMENT OF THE DAMAGE FROM INDIAN OCEAN TSUNAMI ON THE COASTAL MORPHOLOGY OF THE ANDAMAN SEACOAST OF THAILAND BY REMOTE SENSING

THE ASSESSMENT OF THE DAMAGE FROM INDIAN
OCEAN TSUNAMI ON THE COASTAL MORPHOLOGY
OF THE ANDAMAN SEACOAST OF THAILAND BY
REMOTE SENSING
Absornsuda Siripong1
Sombat Yumuang2 and Narumitr Swangphol2

1Marine Science Department, Faculty of Science, Chulalongkorn University, Bangkok, Thailand.
sabsorns@chula.ac.th,
2Geoinformatics for Thailand, Chulalongkorn University, Bangkok, Thailand. ysombat@chula.ac.th
Keywords: Indian Ocean Tsunami, Andaman seacoast, remote sensing, damage assessment, coastline change, Thai coast, NDVI,
coastal morphology, Landsat, runup
ABSTRACT
The Indian Ocean Tsunami that hit countries throughout the Indian Ocean on 26 December 2004 is the most
destructive tsunami ever recorded. It causes the damage on the six provinces of the Andaman seacoast of Thailand.
The degree of the damages were assessed by remote sensing of high and low resolution data with Normalised
Difference Vegetation Index (NDVI) levels. The impact of tsunami wave on coastal morphology at some locations
are exemplified. The relationship of the offshore bathymetry, coastal type, topography, configuration and landuse
with the runup and tide gauge data were combined to explain the cause of different degree of the damage.
THE CAUSE OF TSUNAMI The 26 December 2004 tsunami was generated by earthquake at 00:58:53 UTC
(07:58:53 local time at epicenter). The magnitude 9.15 earthquake near Sumatra is the largest earthquake to occur
since the 1964 Alaska quake and the fourth largest in this century. The main shock ruptured at 2 km/s at the southern
end of the 1300 km long rupture, however, rupture may have propagated northward more slowly. The width of the
fault rupture is 100 km (from the focus up to the bathymetric trench along the megathrust). The average displacement
or slip on the fault plane was 15 meters. The thickness of the fault shear zone across the megathrust is 500 meters.
The whole fault line covers 210,000 km2. A large tsunami triggered by this earthquake spread outward from off the
Sumatran coast. This line source, which oriented nearly N-E, therefore the tsunami directivity was to E-W.
THE STUDY AREA The 6 provinces of Andaman seacoast of Thailand from Ranong to Satul were damaged from
this devastating tsunami. The land use/land cover and coastal environment has undergone severe deteriorated as the
housings, infrastructures, mangroves, sea grass and corals are destroyed in many affected areas. The tsunami caused
significant geomorphologic changes along the coastline, such as eroding sand beaches and enlarging water channels.
The change will impact the ecological balance and adversely affect the life of local people and also tourism.
This will take a long time to recover if unattended or unplanned. There is a need to quickly map the changes and
assess the economic damage. This information is very necessary for the decision makers to plan a strategy for
restoration of ecology, reconstruction and rehabilitation.
METHOD OF STUDY The effect of tsunami on Andaman coast based on the primary appraisal of the effects and
casualty losses on physical and geological conditions of the 6 provinces along the Andaman coast was conducted by
remote sensing technique. We can classify degrees of damage from satellite images into 2 levels depending on data
resolution.
1. Classification effects of tsunami on medium scale
Two Landsat TM imageries (30 meter resolution) used to compute the NDVI (Normalized Vegetation
Difference Index) difference before (March 17, 2004) and after the Tsunami event (December 30, 2004), Fig.1 and 2.
The assumption behind this analysis is that abrupt NDVI reduction in cloud free areas are normally due to a decrease
in vegetation cover or to the presence of water, and can therefore be used as indicators of the potential impact of the
Tsunami. All pixels showing a negative difference were grouped into three classes: “high tsunami impact” if the
difference is bigger than 0.5 NDVI, “moderate tsunami impact” if NDVI is between 0.5 and 0.1 and “low tsunami
impact” if NDVI value between 0 and 0.1.

---

Fig.1. Comparison between tsunami damage assessment and Landsat NDVI Ban Num Khem and Kho Khao Islet.

Fig.2. Landsat TM imageries of pre and post tsunami event during Andaman coast of Thailand (Ko Kho Island to
Tab La Mu, Phang Nga province)Data Source: Landsat TM images courtesy from GISTDA

---

Fig.3. Tsunami damage assessment derived from Landsat TM image along the Andaman Sea Coast
Our analysis indicates area affected by tsunami cover 144.15 sq. km. and classified into 3 levels
(Error! Reference source not found. & 4):
High effects classified by land cover before and after tsunamis are completely change NDVI value
before and after are very different (?NDVI). This type of classified cover 20,265,055.18 sq. km. along the
Andaman coast of Thailand.
Moderate effects classified by different NDVI value are moderate. This type of classified cover
93,154,535.79 sq. km. along the Andaman coast.
Low effects classified by different NDVI value are slightly different cover 30,731,055.32 sq. km. along
the Andaman coast.


Fig.4. Comparison of pre and post tsunami show the complete inundation around Khao Lak beach.

---

2 Classification effects of tsunami on high level of details
High resolution satellite data (e.g. IKONOS, QuickBird) were used to prepare map of pre and post-Tsunami
land use and land cover. Existing maps of land use, land cover, administrative boundaries, mangroves, aquaculture,
plantations and forest will be input into GIS. In addition the socio-economic data will be integrated in GIS. Data
received from different sources will be transformed to one standard and scale for compatibility and analysis. Maps
will be generated at larger scale as the damage is generally located in a narrow strip of approximately 1 Km. All the
generated maps will be supplemented by extensive ground-truth to incorporate attributes related to the damage
including other information. All the data from satellite and other sources will be utilized to develop a map of changes
in coastal ecology, marine ecology, coastal land use (office, residence, hospitals, hotels, etc.). In addition an
assessment will be made in terms of the economic evaluation of the loss in the area.
Study of tsunami effect in high level of satellite imageries especially for the area of Laem Pakarang (Cape
Coral) – Khao Lak beach and area of Baan Nam Khem Villages, Amphur Taguapa, Phang-nga Province, these of
which are the areas that lives and properties were seriously affected most on Thai Andaman Seacoasts.
Damage assessment with high resolution imageries is present in Table 1 and classified into 9 categories:
urban, aquaculture, open area, coconut, orchard, resort and hotel, sandy beach, water bodies and transportation,
definition of each land cover are described in Table 1.

Table 1 Land cover group and definition

Land cover group
Definition of each type of land cover
Urban
Village, town and human resident
Aquaculture
Shrimp pond, and water dike
Coconut
Palm and coconut
Orchard
Rubber, Pine and orchard
Resort
Resort building near coast
Beach
Mud beach, sandy beach
Water bodies
Well, river, stream
Transportation Road,
walkways
Open area
Inundation area, area cover by grass and shrub, sediment deposit by
tsunami wave


2.1 Tsunami Effect on Laem Pakarang – Khao Lak Seashores

Error! Reference source not found.5 show comparing information from IKONOS satellite image, taken
on December 29, 2004 pre- and post- Tsunami events destroying seashore area of Laem Pakarang – Sofitel Magica
Resort and Spa, showing landscape features and geological conditions that have been changed visibly, especially A
section comparing with B section, C section comparing with D section and E section comparing with F section.
In area of A section comparing with B section (the tip of Pakarang cape), This is clear that before the event
of Tsunami, this area belongs to the lush of tropical vegetation (green area). The shape of the tip of Laem Pakarang,
including beach front area, seabed surrounded Laem Pakarang, at present turns to be light-yellow, compounded with
a light-blue, which implies that the sea may have been messed with other substances, which can be sludge and
sediments, together with coral reef deposited, making the tip of the cape extended and the surrounded seabed become
shallow. The area will be susceptible to erosion and wash-away.
The alteration of these areas after Tsunami has replaced the rich vegetation area with sludge and sediments.
The tip of the cape cannot be seen from the image. We have to await the field survey information to support that this
area is under the seawater or the land area. Sued to be above the sea level, it has been eroded and washed away into
the shores by the Tsunami event. This lead to impacts on some seashore areas, namely, seashores have been deeply
eroded as shown by the sea around Laem Pakarang becoming darkened blue, to be concluded that the sea being
penetrated.

In the area of C section comparing with D section (the vicinity of Blue village Pakarang Resort, close to
Laem Pakarang) We can see that before the Tsunami event, the lush tropical vegetation (green area) and a good
conditions of group of building and construction, indicated by an array of orange spot, covers land area. The beach
at the forefront of the resort is visible. The alteration of this area, after Tsunami event, is that the green area
mentioned is piled up with muddy sediments (brown color); the change of the color of sediment in the shrimp
farming area makes the water in the shrimp farm more turbid. Some parts of a group of buildings and construction
has been washed away, orange spots, and that all the beach lines in front of the resort has been eroded and washed
away. All these make the beachfront deeply eroded, together with the apparent widening of a tideway.

---

In the area of E section comparing with F section (the vicinity of Sofitel Magica Laguna and Spa) We can
see, before the Tsunami event, a lush tropical vegetation (green area), a visible beach in front of the resort, a green
vegetation along both sides of the natural waterway (right-hand) and a visible blue pool within the resort. The
alteration of this area, after Tsunami event, evidenced by the replacement of a lush tropical vegetation with muddy
sediments (brown color), an erosion of beachfront, the seashores having been eroded shapelessly, a green vegetation
along both sides of the natural waterway having been pulled out by sea waves making a clearer waterway, a pool
within the resort turning murky, and sediments and mud covering most of the building and construction of the resort.


Fig.5. A map comparing information from IKONOS satellite image, taken on February 11, 2004 (pre tsunami) and
December 29, 2004 (post tsunami) Tsunami events
2.2 Effect of Tsunami on Baan Nam Khem Village

Baan Nam Khem Village, located at Tambon Bang Muang, Amphur Taguapa, Phang-nga Province, is
another area that was destroyed most by Tsunami on December 26, 2004, in Thailand. Baan Nam Khem Village,
located on Pak Klong Nam Khem, which used to be a mining area. Afterwards, a commune of people moved into
this area, had a population of 4,600, comprising of 1,600 families. There also had been laborers from other villages
for an approximation of 1,500 persons and a thousand of Burmese to work for a fishery.

The location and landscape condition of Baan Nam Khem was on the direction of Tsunami attack. There
were no isles or riffles from a distance of the village to reduce the strength of Tsunami upon attack. In addition, by
taking into consideration the depth of sea, Baan Nam Khem location lies in the line of sea trough (like Khao Lak
location), causing Tsunami movement towards the village easily. Due to the reasons mentioned, together with a wide
open landscape of the village, caused a fast and violent Tsunami movement onto the land area, yields a wash-away of
sand dune and a great erosion of seashore and estuary as shown by satellite image produced by IKONOS, comparing
both pre- and post-Tsunami events (Fig.6).

After Tsunami devastated Baan Nam Khem, this area has been altered substantially as evidenced by the
pictures, from the condition of lively villages located at the beach to being the destroyed community with no sign of
lively community left. The classification of primary and secondary effects for this area is very indispensable in order
to plan for zone usage in the future, which may assist in reduction of Tsunami future effect.

---

Fig.6. Pre and post tsunami IKONOS imageries in Ban Nam Kem, Amphoe Takuapa, Phang Nga province.
THE RUNUP DATA The 99 runup transects were surveyed by our team (Fig.7). The highest runup is at Cape Coral
in Phang-nga province for 15.68 meters. The most damaged area is in Phang-nga as shown in Fig.3. Comparison to
the bathymetry and coastal topography, the runups were relatively high at the shallow sea, low-level coastal plain
along the open and sandy coasts. With dense mangroves, coastal vegetation, deep sea, rocky and protected coasts,
the runups were rather low. The configuration of the coastline has some impacts on tsunami runup.

Fig. 7. Runup data.
THE TIDE GAUGE DATA Fig.8. is the result of the analyses of 7 tide gauge station data after filtering the
oceanographic tide. It reveals the characteristics of tsunami waves arriving the Thai coasts (Table 2).

---

Fig.8. Characteristics of tsunami wave from 7 tide gauge data.
Table 2 Time that tsunami waves arrived the 7 tide gauge stations average period of the first 3 waves,
and the arrival times.
Tsunami onset
Average
Time of highest
Tide gauge
time (UTC,
period of the
wave (UTC,
Wave height
Sequence
station
since
1 st 3 waves
since
earthquake)
(min)
earthquake)
Ranong
4:01 (3:02)
53.3
5:20 (4:21)
0.672
First
Kuraburi
3:31 (2:32)
85.3
7:30 (6:31)
0.759
Third
Phuket
2:51 (1:52)
32.5
3:10 (2:11)
0.800
First
Krabi
3:41 (2:42)
63.3
4:50 (3:51)
1.287
First
Kantang
4:56 (3:57)
82.8
5:50 (4:51)
0.780
First
Tarutao
3:31 (2:32)
32.5
4:00 (3:01)
1.073
First
Satun
5:25 (4:26)
68.8
6:37 (5:38)
0.468
First

CONCLUSION The satellite data of medium and high resolutions were used to assess the tsunami damage on the 6
provinces along the Andaman seacoast of Thailand. The NDVI was computed to classify the degree of damage
before and after 26 December 2004. Phang-nga province was the most damaged and Satul was the least damaged
area. The runup data were relatively high at the shallow water, wide coastal plain, sand and exposed beach, while the
runups were relatively low at the deep water, rocky, dense vegetated, and protected coast. The tide gauge data from7
stations were analysed to study the characteristics of the arrival tsunami to Thai coasts. The configuration of the
coastline had some impacts on the runup and damage from tsunami wave. The harbour resonance, funneling effect
and converging phenomena amplified the tsunami waves at Ban Num Khem, Tub Lamu and Cape Coral. The
damage assessment map is a useful guide for rehabilitation and mitigation planning on tsunami disaster.

---

Document Outline

• ??
  • ??

---