GISdevelopment.net ---> AARS ---> ACRS 1990 ---> Oceanography

Barrier Landforms of Thailand - A case study in Narthiwat Province

H.R.H Princess Maha Chakri Sirindhorn
Chitralada Place, Bangkok 10303, Thailand

Dr. Apisit Eiumnoh
INRDM Program, Asian Institute of Technology
P.O Box. 2754 Bangkok 10501, Thailand.


Abstract
The barrier coasts of Thailand can be distinguished on the land sat MSS false color composite (1:1,000 000) of the kingdom. Two sequences of barrier island one in the Bacho didtrict and the other in Mung and Tak Bai districts are found in the Narathiwat province. Particular to this presentation the area in the later case is selected .The land sat thematic Mapper (TM) data acquired on 7 August, 1988 and the aerial photographs scale 1:40,000 taken in 1984 and 1:11,000 in 1988 were interpreted to identify the barrier landform. The study area consists of six barrier islands forming as sand ridges and a lagoon either recent on old between two ridges. The barriers narrow elongated and parallel to the present short line in the northwest and southwest alignment are under coconut cashew tree and fruit-orchard plantations villages are also found scattering on these barriers. The former lagoons are under peat swap forecasts either dense or disturbed ones. Swamp grasses and paddy rice, whereas the recent lagoons have seawater. Because of the different in topography soil moisture and land use types between the barriers and the lagoons they can be delineated on the geometric -corrected land sat TM imagery .The barrier appears on the aerial photographs are used as reference information. It is found that the barrier coastal landform can be delineated best on the land sat TM black and white band 4,5, or 7 for the three band combination false color composite (B,G,R) the very good combination are 2,3,4,: 2,4,7 and 4,5,7 the unsupervised and supervised classification do not provide good information for the barrier study because of the vegetation interferences Consider the cost of the imagery and the availability of the aerial photograph only land sat TM imagery black and white at 1:100, 00-1:50.000 band 4 is suggested for barrier landform mapping in the Narathiwat province.

Introduction.
Barrier coasts consist of a sequence of low ridges of sand built by wave behind the barrier island lies a lagoon which is a broad expanse of shallow water and in many places filled with tidal deposits which then later be covered by native vegetation. Barriers can be developed by expansion of spit bars by the long shore drift and by the up growth of submarine bars. After the regression of the sea level the barriers then found in land parallel to the seashore. These elongated barriers are usually cut by stream to form a tidal inlet and outlet a gap through which strong tidal current flow alternately landward and seaward .The behind lagoon will later be filled layer by layer of marine brackish and fresh water sediments due to the unfavorable reducing condition for plant material decomposition accumulation of peat materials is developed However the barriers consisted of purely quartz sand sediments are cleaned with sorted and fine to very fine sand are well drained condition .The barrier sands are cleaned with very low or nil organic matter content because of the wave action over a long period of time.

From the land sat MSS imagery (1:1.000,000) of Thailand the distributions of the spit barriers and barrier islands are identified in table 1

Table -1 Distribution and barrier types gradient of shelf elevation and aspect of large barrier coasts in Thailand.
District Province Barrier Type Shelf Gradients*
m/km		 Elevations*
m		 AspectAlignment
S. East Coast (Gulf of Thailand)
Muang Rayong Island 10:15 2 E-W
Klaeng Rayong Island 10.25 2 E-W
Tha Mai Chanthaburi Island 10:20 3 NW-SE
Khlung Chanthaburi Island 10:15 3 NW-SE
Munang Trad Island 10:15 3 NW-SE
East Peninsular ( Gulf of Thailand )
Chaiya SuratThani Spit 10:25 3 N-S
Pak Phanang Nakhon Si Thamrat Spit 10:10 3 N-S
Sathing Phra Songkhla Island 10:7 3 NW-SE
Yaring Pattani Spit 10:7 3 NW-SE
Bacho Narathiwat Island 10:4 4 NW-SE
Tak Bai Narathiwat Island 10:4 4 NW-SE
West Peninsular ( Andaman sea)
Thai Muang Phangnga Island 10:1 3 N-S
Langu Satun Island 10:6 3 N-S
*From Topographic Maps, 1969, 1:50,000, Royal: Thai Survey Department.

Thunduan (9184) reported that soils developed on barrier sand are spodosols as indicated by whitish to light gray surface white subsurface and a spodic horizon in the subsoil which is dark brown to dark reddish brown color of accumulated organic materials .The soils fine to very fine sand and excessively well drained are found in humid regions Small areas of spodosols in the southern provinces indicated by Kingdom are Phattalung and Trang provinces.

Okkerman (1983) reported that the most recent transgressive sea level was about ±10,00 year BP. The rise of the sea level was rapid and then slow down to stand still about 3500 year BP in southwest Asia since then transgression and regressions sea level were observed as indicated by marine terraces and barriers Hastings and Liengsakul (1983) mentioned that the Pru to Daeng the peat swamp forest in the study area was dated about 3800 years BP. Pramojanee et al. (1986) reported that the regressive sea level were observed during 4800-1500 years BP. Since then about 1500 years BP the sea level becomes at the present level.

Shao Xiang Ni and Zhen Min He (1984) reported that land sat MSS data with edge enhancement and linear contrast stretching provide good mountain landform mapping but were not good for flat to gentle slope land Jones (1986) mentioned from his study of land sat TM data in arid land that unsupervised classification gives good results for geomorphologic mapping. Other computer assisted are also acceptable for landform mapping units .He recommended land sat TM data band 5 or 7 and unsupervised classification of bands 1,2,3,4,5; 4,5,7; and 1,2,3,4, for geomorphic unit mapping. Ehlers and Boker (1989) stated that barrier island and tidal landforms are. rapidly change .Therefore Land Sat TM data combination of band 1,2,3 was used to detect the change Chaimanee and Tiyapirach (1983) applied aerial photography and satellite imagery for sand bars and beach ridges in songkhla province the northern province of Narathiwat.

Aims
The aim of this study is to find a fast accurate and cheap remote sensing techniques for barriers mapping at medium scale (1:100,000- 1:50,000) in the Narathiwat province. Although aerial photographs arte available but are very restricted to many users . The authors also the disadvantage of the aerial photography and the Land sat TM data.

The Area
The study area located in southernmost province of the Kingdom is between 101o 50'-102o 05 'E longitude and 6o10'-6o 25' N latitude province .The land area is approximately 400 km2 However only barrier sand ridges along the coast are emphasized in this evaluation of barrier mapping unit .The area is in fig 1

Two selected subscences Tak sin (TK) and Saton (KO) each suitable for a computer screen are studied.

Materials and Equipment
The Materials include the following.
  1. Topographic map 1:50,00, 1969 by the Royal Thai survey Department,

  2. Topographic maps 1:25.000 , 1982 by the Royal Thai survey department,

  3. Panchromatic aerial photographs 1:40,000 1984 and 1:11,000 1988.

  4. Land sat TM data geometric corrected dated 7 August 1988 including CCT and hard copies.

  5. The DIMAPS -IBM mainframe and the ERDAS -IBM PC available in the INRDM AIT.
Procedures
This study is the by product of the previous study (H.R.H. Princess Maha Chakri Sirindhorn et.al. 1988) The aerial photographs (1:40,000) are interpreted for barrier and lagoon map units Each band 1,2,3,4,5,7 of land sat TM imaginary 1:250,00 and 1:50,000 are also delineated manually for the same units as indentifed from the aerial photographs .The computer assisted for these landform identification includes displaying of single bands more than two band combination and unsupervised and supervised classifications edge enhancement techniques are employed whenever appropriate .field work was carried out several times and soil profiles were observed on these barriers .The results of barrier and lagoon map units compared with the results from the interpreted from the aerial photograph and a relative value is given.

Results and Discussion.
  1. Interpretation from the aerial photography . The small scale aerial photographs provide very good information for the barrier mapping at the medium scale .The contrast of land features on barrier sand and lagoon is very distinct due to the different in vegetation types land use and soil moisture and characteristics .The barriers consist of coconut and cashew plantations and is wide .The soils are sandy low organic matter and dry .therefore the high reflectance of the electromagnetic energy is recorded on the other hand the langoons havenatural dense vegetation high organic matter and moisture content therefore low reflectance of the wave energy is recorded The smallest dimension of barrier and lagoon is about 50m in width but more than 1kms in length the reief of the barriers obtained from the topographic maps is about 2,4 m the surface slope of these landforms is very gentle to very slightly undulating.

    The large scale aerial photographs can be used for this landform mapping but it requires may photographs for the work therefore they are used only ae reference whenever necessary.

  2. Single band of the Land sat TM imagery Since it is very time consuming to work on the computer screen photographs of each display were taken directly from the computer . Photographs of each land sat Tm band were then compared with the results from the aerial photograph interpretation The land sat TM bands4,5 and 7 give good results comparable to the aerial photographs It may be noted also that 4 is probably the best.

    The photographic products enlarge from each of the Landsat TM which were plotted on 22x22 cm film (1:500,000) gave similar results The Land Sat TM banda 4 and 5 are the best (table 2)

    The results confirm the appropriate wavelength and the ground resolution designed for the land sat 4 and 5.

  3. band combinations of the Land sat TM imagery the results of three band combination (B-G-R) are in table-3 combinations of 2,3,4 :2,4,7: and 4,5,7 give better results than the other combinations .

  4. Unsupervised classifications .The unsupervised classification of both subscenes by using various band combination were not satisfactory for barrier mapping .Similar results from the supervised classigfication were also obtained . Both classifications were not suitable for barrier mapping due.

Table -2 Comparison of barrier distinctness as seen from each band of
the Land sat TM data ( 7 August 1988 ) of the Taksin area (TK) and the KO Sathon area (Ko)
Band Computer Display* Photo Plotter**
  Tk Ko Tk Ko
1
2
3
4
5
7		 D
C
C
B
B
B		 D
C
C
A
A
B		 C
C
B
A
A
B		 D
D
C
A
A
B

*512x512 pixel-line of the geometrically corrected Landsat TM dta displayed on the ERDAS PC.

** Geometric corrected TM data with Lin enhancement: Histogram then plotted on 22x22 cm film band by band (1:50,000) before printing at large scales.

Note :   A- Very Good       C= Fair E= very Poor
             B- Good               D= Poor

Table -3 Comparison or barrier distinctness as seen from each combination of
land sat TM data (7 August 1988) of the Tak Sin (TK) and the Ko sathon areas
Bands (BGR) TK KO
247 A A
154 B A
432 B D
321 C D
457 A A
123 B D
451 C C
741 C B
751 B B
235 - B
234 A A
345 - B
374 B -
753 B --

Note : 1) Each Sub area is 512x512 pixel line of the geometrically corrected landsat Tm data displayed on the ERDAS PC. 2) Relative comparison : A =very Good , C= Fair D= Poor and E= very poor.

Reference.
  1. Chaimanee N. and s. Tiyapirach 1983 on the coastal morphology songkh;a province southern Thai-land pp. 80-95 In Geomorphology and quaternary Geology of Thailand .Proc. of the first symposium Bangkok. October 1983 Department of Geology Chulalongkorn Univ Dept of mineral Resources and Goel.

  2. Ehlers J. and F Boker 1989 Detection of short term changes in tidal flat geomorphology .pp. 149-159 In proc. Of the Workshop on Earthnet Pilot Project on Landsat Thematic Mapper Application December 1987 Frascati Italy.

  3. H.R.H Princess Maha Chakri Sirindhorn S. Silapaacharanan and K. Srisukasawadi 1988 Land use / Land cover map accuracy assessment of Land sat thematic Mapper data using the DIMAPS image processing system for Narathiwat province . pp. 1-15 In proc of the 9th Asian Conference on remote sensing November 1988Bangkok Thailand.

  4. Hastings . P. and M liengsakul 1983 -Chronology of the late Quaternary climatic changes in Thailand .pp 24-34 In Geomorphology and Quaternary geology of Thailand Proc of the first Symposium Bangkokoctober1983 Department of geology Chulalingkorn Univ Dept. of mineral Resources of Goel Soc. Of Thailand.

  5. Jones A.R 1986 Unsupervised classification of thematic Mapper imagery for geomorphologic and sediment logy pp. 64-79 In geomorphologic investigations pp. 647-657 In proc of a symposium on Mapping from modern imagery September 1986 ISPRS Edinburgh Scotland.

  6. Okkerman , J.A 1983 Coastal geomorphology and sediment logy pp. 64-79 In geomorphology and quaternary Geology of Thailand Proc of the first symposium Bangkok october1983 Department of Geology Chulalongkorn Univ dept of mineral resources and Goel soc Thailand .

  7. Pramojaneee P,. P Hastings M, Liengsakul and Engkul 1986 TheHolecene transgression in peninsular Thailand pp. 551-564 In GEOSEA V Proc. Goel SOC Malaysia April 1986 Goel Soc. Malaysia .

  8. Shao Xiang Ni and Zhen Min he 1984 Digital Land Sat Image enhancement for geomorphological mapping of the southern Huairon county ofBejing China pp 421-425 Proc. Of remote Sensing Views and Previews September 1984 Reading Berks UK.

  9. Thunduan V. 1984 Properties genesis and distribution of Podzols (spodosolaThai land Ph. D Thesis Kaset sart Univ Bangkok Thai Land 325 p.