Evaluating Mangrove Extent Mapping from Landsat 8 and Sentinel-2 Using the Minimum Distance Classifier Algorithm
Document Types
Poster Presentation
School Name
De La Salle University Laguna Campus
Track or Strand
Science, Technology, Engineering, and Mathematics (STEM)
Research Advisor (Last Name, First Name, Middle Initial)
Masongsong, Angela Nicole, S.
Start Date
23-6-2025 10:30 AM
End Date
23-6-2025 12:00 PM
Zoom Link/ Room Assignment
5th Floor Breakout Function Room (501-503), Enrique K. Razon Jr. Hall, DLSU Laguna Campus
Abstract/Executive Summary
Mangrove cover in the Philippines decreased by more than 50% by 2000, from an estimated 400,000 to 500,000 hectares in the 1920s. Mangroves play an important role in preserving species biodiversity, allowing carbon sequestration and coastal resilience. Monitoring the condition of mangrove forests is essential because of the serious threat posed by deforestation for road building and aquaculture conversion. By offering a cost-effective and efficient way to map vast areas, satellite imagery has transformed ecological research and made it possible to identify and monitor mangrove forests both locally and globally. This study aims to assess the effectiveness of Landsat 8 and Sentinel-2 satellite imagery in classifying and mapping mangrove cover through the means of validating the accuracy by conducting supervised and unsupervised classification algorithms. The study used both unsupervised classification through clustering and supervised classification with the minimum distance classifier through the Semi-Automatic Classification Plugin in the Quantum Geographic Information System (QGIS) software. The results show that clustering output for unsupervised classification received a total of 44% total accuracy for Sentinel-2 and 38% total accuracy for Landsat 8. In the case of supervised classification using the minimum distance classifier, Sentinel-2 has a 74% total accuracy, while Landsat 8 resulted in a total accuracy of 70%. The study concludes that supervised classification performed better compared to unsupervised classification based on the accuracy assessment using a confusion matrix. This study contributes insights into the application of open-access satellite imagery to mapping mangroves.
Keywords
mangrove monitoring; remote sensing; Landsat 8; Sentinel-2; image classification
Research Theme (for Paper Presentation and Poster Presentation submissions only)
Sustainability, Environment, and Energy (SEE)
Initial Consent for Publication
yes
Statement of Originality
yes
Evaluating Mangrove Extent Mapping from Landsat 8 and Sentinel-2 Using the Minimum Distance Classifier Algorithm
Mangrove cover in the Philippines decreased by more than 50% by 2000, from an estimated 400,000 to 500,000 hectares in the 1920s. Mangroves play an important role in preserving species biodiversity, allowing carbon sequestration and coastal resilience. Monitoring the condition of mangrove forests is essential because of the serious threat posed by deforestation for road building and aquaculture conversion. By offering a cost-effective and efficient way to map vast areas, satellite imagery has transformed ecological research and made it possible to identify and monitor mangrove forests both locally and globally. This study aims to assess the effectiveness of Landsat 8 and Sentinel-2 satellite imagery in classifying and mapping mangrove cover through the means of validating the accuracy by conducting supervised and unsupervised classification algorithms. The study used both unsupervised classification through clustering and supervised classification with the minimum distance classifier through the Semi-Automatic Classification Plugin in the Quantum Geographic Information System (QGIS) software. The results show that clustering output for unsupervised classification received a total of 44% total accuracy for Sentinel-2 and 38% total accuracy for Landsat 8. In the case of supervised classification using the minimum distance classifier, Sentinel-2 has a 74% total accuracy, while Landsat 8 resulted in a total accuracy of 70%. The study concludes that supervised classification performed better compared to unsupervised classification based on the accuracy assessment using a confusion matrix. This study contributes insights into the application of open-access satellite imagery to mapping mangroves.
https://animorepository.dlsu.edu.ph/conf_shsrescon/2025/poster_see/3
