Study of land cover in “Lanna bioregion”: a case study of Suan Sak campus, Chiang Mai University

Abstract

This article aims to classify land cover from satellite imagery using Geographic Information System (GIS) with a supervised land cover classification method in the study area of Suan Sak Park, Chiang Mai University. The study compares the results with data obtained from field surveys under the Chiang Mai University’s Green University Ranking 2019 project. The study found that the overall accuracy of the supervised land cover classification in the study area was 83.00%. The Kappa coefficient was 0.74. It was found that the user’s accuracy in the water area was the highest at 100%, while the lowest accuracy was in the forest class at 66.67%. When comparing the results of land cover classification with the field survey data, a significant discrepancy in the quantity of land cover was observed. This discrepancy was mainly due to the low spatial resolution of the satellite imagery used for analysis, which limited the classification of land cover in small areas, particularly small-sized gardens and narrow waterways. However, applying the mentioned land cover classification method in conjunction with field surveys not only improves the accuracy and precision of the data but also reduces the time and cost of on-site surveys.