The remote and inaccessible nature of many tropical forest and the rapid changes in
land use/land cover make remote sensing techniques an essential tool to monitor the
landscape and generate valuable data, which can be interpreted by various technique
and methods. Multi-temporal and multi-spectral data from Landsat 5 TM after
image analysis successfully delineated the land use/land cover changes in Ban Don
Bay, Surat Thani, Thailand. Most of the changes were observed in the wastelands,
shrimp farms, and vegetation. From the field visit it was found that several changes
are due to the development of plantations: rubber and coconut. It was also found
that the majority of the changes have occurred during 1993–1996 when compared to
1990–1993 and 1996–1999. Local people developed shrimp farms on a large scale
during 1994 by converting available wasteland, and vegetation around coastal areas because of its commercial value. Because of strict Thai regulation, mangrove and
forest was not allowed to cut, and it also grew. An increase in the forest, mangrove,
urban and shrimp farms was observed in figure 4. This increase is at the expense of a decrease in the wastelands and agricultural area. Generally shrimp farms adversely
affect the mangroves, but thanks to strict Thai regulations mangrove cover has
increased.
Changes observed in figures 6 and 7 are obtained using NDVI. Results from this
approach also indicate an increasing trend in mangrove coverage. The NDVI
differencing approach mostly highlights changes from vegetation to other land use
or from other land use to vegetation. Land use changes from one category to
another are shown clearly in the NDVI composite image in figure 8. K-mean
classification of the NDVI composite image was found useful for delineating the
development of shrimp farms in the bay (figure 9). The methods employed in the
study and Landsat data were useful to investigate the land use change pattern in the
Ban Don Bay, but the results would have presented more details if high spatial
resolution data were utilized. Remote Sensing demonstrated the potential for
monitoring the land use resources. Monitoring and planning of land use in coastal
zones can be effectively carried out by integrating and analysing the output raster
maps in GIS.