GISdevelopment.net ---> AARS ---> ACRS 1990 ---> Geology Disaster

An overall application of Remote Sensing and GIS for disaster assessment: A case study for Chumphon Province

Dr. Suvit Vibulsresth,Dr. Darasri Downreang
Supak Wongparn,Ramphing Simking
Remote Sensing Division, National Research Council
Bangkok 10900, Thailand


Abstract
The Gay typhoon had shipped across the Gulf Thailand on November 3, 1989, terrorizing fisherman and sinking hundred of fishing vessels before it swept through the southern part of Thailand, flattering villages and agricultural land as it cut its destructive path across the isthmus.

An attempt is made to integrate remote sensing technology with geographic information system available at the remote sensing division of the national research council of Thailand in order to obtain information that could be useful and needed by the central government for establishing a plan for recovering the damaged area. This paper describes how the data are captured, selectively retrieved, analyzed for the assessment of disaster impacts related to sudden environment and landuse changes.

Introduction
The "GAY" typhoon swept across Chumphon province which is located in the southern region of Thailand, in the beginning of damages to property and agricultural area. Therefore, after the tragic situation, the government has immediately set up programmes for rehabilitation and relief. This report was prepared by RSD/NRCT in order to support the said programmes. This paper emphasized on the damages that occurred in area of economic corps via SPOT image processing technique. The results were then integrated with other information, for example, soil types, road network and elevation contours, by using geographic information system in order to make an assessment of disaster geographic impacts related to sudden environment and land use changes.

Objectives
The main objectives of the study are as follows:
  1. To use satellite data to map the existing resources in the disaster area. The features to be mapped include:

    • the damaged economic crops

    • the land use pattern before and after the Gay,

    • water resources, and

    • the present road network.

  2. To use the GIS in building up a database of the area. This database will incorporate the maps derived from satellite data in (2.1) and maps from other sources for information such a elevation contours and administrative boundaries.

  3. To apply the information contained in the database in the forecast for soil loss potential and its impacts on the environment.

The Gay Typhoon
Before this unfortunate situation, since October 30m 1989, the Meteorological Department of the Ministry of Transport and Communications had reported daily the weather forecasts on the formation of the typhoon called "Gay" from eh low pressure originated in the lower part of the Gulf of Thailand. About 5 days later, on November 4, the Gay struck the coastal are of Amphoe Pathiu and amphoe tha-sae. It then moved across the Kra Isthmus tot eh Andaman sea and finally became a depression on November 5.

The report of the adhoc committee for Rehabilitation and Relief of the "Gay" typhoon revelaed that 32, 301 families became the victims of the Gay. The Disaster in Chumporn province covered an are of 4,605 square kilometers including Amphoe Pathiu, amphoe Tha-sae, Amphoe Muang Chumphon, Amphoe Kraburi and Amphoe Bang-Sa-Phan. However, amphoe Pathiu, dn Amphoe Tha-Sae were considered the most severally damaged area since they were in the center of the movement of the Gay.

Applications of Remote Sensing and GIS
  1. Determination of the Gay affected area using GIS/SPANS

    It is essential to accurately define the are under the destruction of the Gay for this study. Too large the are would be time consuming whereas the smaller area might not be sufficient for a sound conclusion and recommendation to be made. The following steps therefore were taken to identify the disaster area.

    1. The track of Gay and its radius of 50 km (0n 4th November) as reported by the Meteorological Department were entered into the GIS/SPANS to define the coverage of the damage area which extended 50 km from bothsides of the center lien of movement. The resulting damaged zone when overlaid on 1:50,000 map index is as shown in figure 1.


      Figure 1 The damaged zone overlay on 1:50,000 map index.

    2. The administrative boundaries at a district level were digitized from1:50,000 topographic maps for the coverage zone in described (4.1.1). the derived map showing district, provincial and national boundaries is shown in figure 2.


      Figure 2 The administrative boundaries map.

    3. The damages zone was subdivide further into 2 levels according to the information obtained from the preliminary ground survey conducted between 23-26 November 1989. The first level is the zone extending 30km from the center line of movement where the damages were between 50-100%, and the second is between 30-50 km with lesser damages of 0-50%.

    4. By overlaying the administrative boundary map and the damaged zone map with GIS/SPANS, the extent of damages in districts under the sweep of Gay could be assessed (as shown in figure 3). It was found that the total damaged are within the 50 km radius is about 4,605.4 sq.km and the at Amphoe Pathiu and Amphoe tha-sae in Chumporn province were most affected with 50-100% damages.


    Figure 3 The extent of damages in districts under the sweep of GAY

  2. Assessment of the damaged economic crops

    In this part, remotely sensed data were used to identify and assess the economic crops that had been damaged by the typhoon.

    1. Ground pre-survey: Observation of ground condition was made between 23-26 November 1989 to examine the nature and extent of the damages especially those of economic crops. It was found that the severely damaged area is Amphoe Pathiu and Amphoe Thasae as indicated in(4.1). this is the area under the coverage of 1:50,000 map sheet no. 4830 I, 4830 II, 4830 III and 4830 Iv ad shown in figure 1. The crops that were mostly destroyed were coconuts, rubbers, oil palms and orchards while paddy fields were only partially affected from floods. These area were selected to define training classes in the satellite data analysis.

    2. Satellite data analysis on the image analysis system Meridian; The information obtained from ground observation was used to identify from satellite data the damaged cropping area in the two districts covering 4 map sheets mentioned above. Only two main crops, i.e para rubbers and oil palms, would be classified due to the fact that the size of their plantation plot's was large enough to be accurately determined. As for coconuts, the plantation pattern was too irregular with less crown density for classification to be made with the same level of accuracy as the other two crops, and therefore was not included in the analysis.

      CCTs of geocoded SPOT multispectral images were used for the analysis. These images were taken before and after the typhoon, on 13 December 1988 and 23 December 1989 respectively. From the enhanced false colour composite images, rubber and oil palm plantations could be distinguished from their colour in representation and its changes after the disaster, as described in table 1. The changes of colour were due to the falling of the trees and the dying leaves that no longer contained chlorophyll substance.

      The images of both period were then superimposed for the purpose of defining training areas of the damaged crops. Digital classification was finally made to produce a map showing the damaged rubber and oil palm areas.

      Table 1
      Crop type Before disaster
      on 13 December 1988      		 After disaster
      on 23 December 1989
      Rubbers oil palms Dark red and pinkish red Greenish blue light red

    3. Transfer of satellite derived maps to GIS/SPANS; The classification results of the damaged crops were directly transferred from the Meridian to the SPANS data base. In addition, the road network that was updated on the 1:50,000 base maps was digitized into the same data base. The updating was done through interpretation of geocode SPOT panchromatic images (12 November 1988 and 8 February 1989) using the PROCOM-2 image projector system. The overlay of the damaged crops with administrative boundaries and road network is shown in figure 4 as an example for Amphoe Pathiu.


      Figure 4 The damaged crops overlay with administrative boundaries and road network, Amphoe Pathiu only.

    4. Application of the GIS

      In view that the damages had left the area with barren soil which would be prone to erosion, the exercise of this study was geared therefore towards the use of GIS to forecast the area of soil loss potential. The data necessary for this study are those already described including the soil maps which wee also digitized into the data base. These soil maps were published at 1:100,000 scale by the Land Development Department.

      As indicated in the soil property table that the surface runoff is dependent on the soil profile, slope, climate and vegetation cover, the surface runoff therefore can be used as a key factory to indicate soil erosion potential with a relationship as described in Table 2. the fact that the mean annual rainfall of Chumporn is as high as 2,000 mm with possibility of short breaks of heavy rain due to the prevailing monsoon, it surface runoff and vegetation cover, based on two parameters namely, surface runoff and vegetation cover, under the assumption that during rain, areas prone to erosion are those without land cover and with high surface runoff.

      Table 2
      Surface runoff Soil erosion
      Slow None to little
      Medium Slight to moderate
      Rapid Moderate to high

      In order to create the surface runoff map, the soil series maps ere first digitized into the data base (eg. Figure 5 for Amphoe Pathiu). Then attribute data were constructed to relate principal soil properties with soil series units. A new modified soil unit map was finally generated in terms of surface runoff as illustrated in figure 6 (Pathiu area).


      Figure 5 The soil series map, Amphoe Pathiu only.


      Figure 6 The surface runoff map, Amphoe Pathiu only.

      By overlaying the maps of damaged corps and surface runoff and using the relationships mentioned earlier, the erosion potential map could be produced such as presented in figure 7.

Figure 7 The erosion potential map, Amohoe Pathiu only.

Result/Conclusions
  1. Forecasts of future possibilities

    With the sudden changes of land environment particularly in Pathiu and Thasae and the inclining topography from the mountain range in the West to the sea in the East, the following might occur in the next rainy season.

    1. The overflowing runoff in Klong thsase and upper Chumporn river could cause flooding in low land along river course sources and north of the city.

    2. The eroded soil could produce shoaling effect of the rivers well the poor quality of water.

    3. The loss of surface soil would deteriorate the soil fertility and make in unsuitable for cultivation.

  2. Recommendations

    The information obtained from the study has helped understand the land condition of the area. The following area some recommendations for rehabilitation and relief of the disaster area.

    1. The remaining resources including land and water should be first protected and revived as soon as possible. This could be done by means of crop diversification. These crops should be fast growing type with spreading root system. Moreover, they should be the plants that can grow under bad soil as well as can be used to feed animals.

    2. Rehabilitation activities should be include the upgrade of people's living conditions through promotion of farming exercises both for short term and medium term programmes. The idea is to use the diversified crops for feeding purpose and at the same time the manure obtained from the farms can be used for cultivation. In the long term plan, this area can also be promoted to be a regional communication, centre as well as a tourism spot.