Change Detection in Coastal Zone by Remote Sensing Technique

 

Mrs. Valairat Wanpiyarat, Ms.Promchit Trakuldist, Mr. Boonrak Pattanakanok
Land Development Department,
Ministry of Agriculture and Cooperatives, Thailand.
http://www.gisdevelopment.net/aars/acrs/1995/ps1/ps1012pf.htm

Abstract
Because land use information was limited, therefore, the Landsat TM color composites, band 432 (ROB) acquired on May 29,1989, and band 453 (ROB) acquired on September 24,1991, and September 10, 1992, were visually interpreted for land use changes detection in to the coastal area of 340 km 2 at Amphoe Pak Panang, Nakhon Sri Thammarat province. The results of interpretation and field verification revealed that the increasing of shrimp fans , during 2 the past three years, was up to 25.78 km due to the encroachment of mangrove forest and land use change from paddy land to shrimp fanning. Dispersed salt water, generated from shrimp fans, created soil salinity and water contamination to the nearby paddy fields, resulted in paddy abandonment. The total abandoned paddy field was found to be 17.80 kIn The coconut and banana area increased about 9.84 kIn2. Approximately 13.63 kIn2 of the mangrove forest was destroyed and the highest depletion area, 9.72 kIn 2 ,was found between the year 1991 and 1992.

In addition, the Landsat TM images also showed that the coastal shore lines had been changed from year to year indicating the usefulness of satellite image for update mapping.

I. Introduction
The history of development of the Pak Phanang River Basin has long been more than 10 years. The first real developmeilt concept was created in 1980, with consideration of the Large-

Scale Project Section, the Project Planning Division, Royal Irrigation Development. Up to now the serious problems in the Pak Phanang River Basin can be described as

  1. Shortage of fresh water for irrigation development and for domestic uses,
  2. Salt water intrusion into the Pak Phanang River and tributaries which lasts about 9 months each "year, causing shortages of fresh water into the lower Pak Phanang basin,
  3. Conflicts of interest between rice farmers and marine shrimp farmers due to discharges of salt water from shrimp ponds into paddy fields, causing damages of rice fields along the coastline of the Gulf of Thailand, along the Pak Phanang River, and along the Khlong Cha Uad,
  4. Annual flooding problems which usually occur in the period from November and December reach year causing damages of houses, infrastructure, and agricultural areas of more than 1 ,248 km

These serious problems have to be mitigated, and this is the reason why it is necessary to develop Pak Phanang River Basin.. His Majesty the King, during his visits to Southern Thailand m 1992, gave the Royal Irrigation Deportment his concept of building salt water barrage tat Ban Bung Pi, Tambon Hulong, Amphoe Pak Phanang, Nakhon Sri Thammarat province as well as a series of drainage canals in order to solve the above -mentioned problems

In order to mitigate the conflicts between the rice farmers and the shrimp farmers in the area, land use zoning delineation between fresh water and saline water zone is very necessary Therefore, the Department of Land Development. has carried out land use surveying, emphasizing on land use changes in the paddy area and the mangrove forest. Those existing land uses including the distribution of shrimp farms are basically required for creation of reasonable dividing line.

2. Study Area and Objective
The study area is located between 8 ° 15 to 8 30 N latitudes and 100 05 to 100 17' E longitudes covering about 340 km2 in the Pak Phanang River Basin , eastern side of Nakhon Sri Thammarat province of the southern peninsula Thailand. Its physiography composes of bays,Li sandy beaches, river mouth, dunes, and tidal flat. The north side of the study area in the river mouth of the Pak Phanang River, is covered with mangrove forest while the rest mostly contains agricultural land ( Figure 1 ).


a. May 29, 1989. Band 432

 


b. September 24, 1991. Band 453

 


c. September 10, 1992. Band 453

Figure 1. Landsat images of the study area, acquiring on May 29, 1989, September 24, 1991, and September 10, 1992 respectively.


Remote Sensing technology provides multiple, repetitive, and update physical information. For this research, the authors aim to apply remote sensing technology to monitor land cover and land use changes especially from mangrove forest and or paddy to shrimp farm, and also estimate shrimp fanning area in the study area.

3. Methodology
The methodology includes an analysis of temporal Landsat TM images, field verification, polygon overlay, and monitoring of land use changes.

An amount of three Landsat TM images, acquired on MultiMate between the year 1989 and 1992 and produced at scale 1 : 50,000 was visually interpreted for land cover / land use investigation. The qualitative comparison of land use changes work during d1e period of d1ese data.

To observe the circumstances, field verification was carried out in May 1992 and September 1993. In the field, land cover land use, forest condition, soil and water condition were described and recorded.

To study land cover land use changes, the multiple layers of previous interpretation maps were overlaid to create new polygons showing those changing areas. A number of changed polygons were examined and calculated for reporting.

4. Result and Discussion

Analysis of Land Cover /Land Use
Land use categories are distinguished on the color composite of Landsat TM band 234 and 354 ( BGR ), based on differences in color and texture. Those categories are recognized as : paddy, mixed fruit tree, shrimp farm, built up land, mangrove forest, marshland, and others. Paddy and mixed fruit tree are the major agricultural crops in the area. Both irrigated and non-irrigated paddy fields cover about 49 % of the total study area , 340 kIn 2. Mixed fruit trees area, is; coconut, banana, and orange, were found to be approximately 7% in the year 1989. To identify the type of fruit trees from Landsat TM images without field checking is impossible. Active shrimp farming areas, along the coast lines, along the Pak Phanang River, and in mangrove forest, are seen clearly from both combinations of Landsat TM. Dense mangrove forest can be recognized as bright red color with fine texture from the 234 combination and as peach color from the 354 combination. Amphoe Pak phanang is shown in bright blue color along both sides of the Pak Phanang River which flows downward from the river mouth to the south. The other land uses are i ~ marshland and abandoned paddy which can be identified by utilizing temporal data.

Land Use Change Analysis
An analysis of mechanisms of land use changes plays an important role in forecasting environmental changes. As the population increases, infrastructures develop, and economy blooms in the study area and its vicinity , these cause changes in the coastal environment and directly effect on the benefits of the farmers. Hence, we consider the factors of land use changes in the study area from the view point of human impact.

During the three years observed period, the major land use changes are as follow:

  1. Changes from paddy to shrimp farming,
  2. Changes from mangrove forest to shrimp farming,
  3. Changes from paddy to abandoned paddy.

Figure 1 shows all of those changes from the year 1989 to 1992. Results from visual interpretation and field observation revealed that, during the year 1989 to 1991, paddy field had changed to be shrimp farm at the changing rate of 9.1 % and this change had slightly decreased to be 7.1 % during the year 1991 -1992. Shrimp farming expanded dramatically, at the increasing rate of 116 % .in the year 1991, from mangrove forest particularly in the north side of Amphoe Pak Phanang, to nearby paddy and to paddy along the coastal zone. The shrimp farming area was found to be 40.43 kIn2 in the 1992 image while it was only 14.6 kIn2 in 1989. Salt water, generated from shrimp farms, discharged into paddy fields causing soil salinity and water contamination. These problems had affected seriously to rice farmers by lowering rice yield to a very minimum, therefore some of paddy fields were left abandon. The abandoned paddy area increased from 6.1 kIn in the year 1989 to 17.89 kIn in the year 1992. In the paddy fields and along stripped villages where were not affected by salt water, farmers planted fruit trees, coconut and banana, and this area rapidly increased approximately 9.84 km 2 within three years.

In spite of forest plantation at Laem Pak Phanang, mangrove deforestation still took place at due alarming rate due to shrimp farming. The destroyed mangrove forest area was found approximately 13.63 km2 The highest depletion occurred between due year 1991 and 1992 with the rate of 8.62 % resulted in 9.72 km 2 loss.

In addition to due coastal land use changes, due extent of shore lines changes could be observed from due multi-temporal Landsat TM images. From due comparison among these images and information obtained from topographic maps, the findings could be summarized as follows :

  1. The shore lines at the Pak Phanang River Mouth and along the Pak Phanang Bay advanced toward the sea due to forest plantation.
  2. The north spit at Laem Talumphuk extended in a north -west direction to some extent ( Figure 1 ).

5. Conclusion
Coastal land use changes can be monitored using overlay monitored of multi-temporal Landsat images. During the observed period of 1989,1991, and 1992, the major changes have been observed as changes from paddy and mangrove forest to shrimp farming as well as abandoned land. Rates of these changes are determined and total shrimp farming area is calculated. Shorelines movement is also detected from the study.

6. Reference

  • F.A.O. 1982. Management and Utilization of Mangrove in Asia and Pacific Region. Environment Paper #3. Rome. F.A.O. ;
  • F.A.O. 1984. Mangrove forests in Asia Pacific Region: A Summary of Available Information. Bangkok. F .A.O.
  • Pheng, K.S. and Foo L.K. 1989. Detection of Land Use Changes Using Remote Sensing and Proceedings of the Tenth Asian Conference on Remote Sensing. November 23- 29,1989. Kuala Lumpur, Malaysia.