Forest encroachment mapping in Cameron Highlands, using cadastral parcel and remote sensing datasets

Abstract

High demand of agriculture product from Cameron Highlands has caused unauthorized agriculture activities to emerge. When agricultural land is highly in demand and agriculture land becomes scarce, the government land, mostly forest areas are being encroached for this purpose. The forest encroachment which include cutting down trees on hill slopes, can cause negative impact on forested area in Cameron Highlands including Mentigi Forest Reserved (MFR) area, which also serve as water catchment area. This study is to detect and assess this encroachment especially that caused by agricultural activity in MFR area and its surrounding, using remote sensing datasets and cadastral parcel. The hybrid classification method was applied on remote sensing images to classify the land-cover types in the study area. Cadastral parcel was used as a reference for forest reserve area boundary to detect some agricultural activities inside the forest reserve area. The findings reveal that around 8.75 ha of forest encroachment was detected using SPOT 5 image of year 2010, and 2.98 ha were detected using IKONOS image of year 2001. It was found that most of the encroachments were meant for agricultural purpose. Accuracy of the extracted forest boundaries from satellite images was assessed using the Root Mean Square Errors (RMSE). The displacement value for IKONOS and SPOT 5 were 8.11 m and 12.99 m respectively. These errors were further reduced using buffer zones that was applied to reduce the error effect on encroachment detection. This method was also applied to the entire subset of satellites image to identify encroachment areas inside the forest reserve and surrounding areas. Results of this analysis show that forest encroachment occurred inside MFR area despite its relatively small size. The study has proven that the technique used is an effective method in assessing small forest encroachment area.