Development of forest encroachment model from satellite image in Nan Province.

摘要

Encroachment of forest areas in Nan province has continuously increased. Although there are many agencies responsible for monitoring and protecting natural resources including forest areas. Due to the number of forest ranger is rather limited and not adequate to monitor and protect all forest areas. Therefore, forest encroachment model was created using satellite image in Nan Province in order to support forest ranger. The objectives are to design and develop a model to analyse forest encroachment sites accurately. The model used LANDSAT-8 data between the two periods of time (Before and After). Clouds and shadows were erased from the images to analyse the difference of the NDVI between the two periods of time. Then the data was reclassified using Unsupervised Classification's hereby the data groups with has high difference values in forest areas and other conservation areas that larger than 1 Rai (0.0016. Sq.km.) were selected. Then the data was converted into points. In the next step of the model, the point data was filtered in accordance with associated and geospatial information for example Areas for permission to use from the government etc. Visual analysis of satellite images between the two periods of time was also made and the comparison with aerial and satellite images 7 years retrospective. The field survey was proceeded to inspect at the actual areas. The results of the analysis using the forest encroachment model from 20 points It was from that 16 points of forest encroachment were detected. While 4 points were not as a forest encroachment, where are shifting cultivation areas. The percentage of accuracy of the model is 80% . Improvement of the model can use satellite images retrospective for many years and GIS data for the government related to agriculture in Nan province. For the most precision and accuracy of point data.