A thesis submitted in partial fulfillment of the requirements for t he degree of Master of Science. School of Environment, Resources and Development

Thesis (M.Sc.) - Asian Institute of Technology, 1986

Landsat MSS data were used to assess the feasibility of utilizing for identification of different land use/cover types and their multitemporal distribution in the Eastern Part of Chao Phraya Plain, Thailand, using both visual interpretation (manual) and computer assisted analyses. The results of this study show that black/white imagery of MSS band 2 and 4 and false color composites in single or in combination with other bands were found appropriate to identify the different land use/cover types from visual interpretation, and band 4 was found excellent in determining land use/cover types and soil properties such as soil moisture contents, soil textural composition, soil organic matter contents and soil color from computer analysis. Ten land use/cover types such as evergreen forest, mixed deciduous/ evergreen forest, scattered plants/scrubs/grasses, perennial crops, paddy, fallow land of dry/light color soil, fallow land (wet/dark color soil), burned land, urban land and water body/swamps were identified from both visual and computer analyses during dry season. On the other hand, only seven land use/cover types such as evergreen forest, mixed deciduous/evergreen forest, scattered plants/scrubs/ grasses, perennial crops, paddy, urban land and water body/swamps were able to identify from visual analysis. 57 percent of the total land in the study area was found fallow during dry season of the year. The results indicated that the visual interpretation (manual) provides a simple and inexpensive approach for analyzing Landsat data but whenever considering the accuracy and detail of the mapping unit , the computer analysis has found superior results. 79 percent accuracy was observed from computer aided analysis of Landsat multispectral data as compared to 60% and 47% from visual analysis for dry and wet seasons respectively. Computer aided analysis of Landsat MSS data yields highly satisfactory results in spite of complex land use/cover environment in the study area. Therefore, visual interpretation and coinputer aided analysis of remote sensing MSS data should be used complementarily and consideration must be given to which approach (or combination of techniques) of remote sensing best fits a particular application for land use/cover types and other resource management inventories.