Philippine E-Journals

Land Cover Classification of the Abra River Basin with Remote Sensing and Machine Learning Algorithms

NOVER MATSO

 

Abstract:

Land cover is the biophysical or the visible cover on the Earth’s surface. It may also refer to the man-made and natural characteristics of the Earth’s surface, such as natural vegetation, soil, agriculture, and human infrastructure. This study was conducted to classify the land cover of the Abra River Basin with the use of GIS and Remote Sensing Technology and machine learning algorithms such as Support Vector Machine and Random Forest. The river basin was classified into six broad classes: agriculture, grassland, built-up, water bare ground, and vegetation. The dominant land cover class of the Abra River Basin is vegetation followed by grassland, agriculture, built-up, bare ground, and water. The support vector machine classier produced the highest accuracy. The result of this study is recommended to be utilized in the conduct of other research such as geo-hazard mapping, ecological risk index assessment, and species distribution modeling. The land cover map produced can be used as input in preparing a watershed management plan for each of the municipalities covered by the Abra river basin.