Spatiotemporal Pattern of Mangrove Forests in Bulacan Using Satellite Data Fusion Method and Machine Learning

• Main Authors: Aperocho, Aaron Anthony, Chan, Stephen Raphael, Mangali, Juan Miguel, Orpeza, Richmond James Marthy, Purio, Mark Angelo
• Format: Conference Proceeding
• Language: English
• Subjects: Machine learning algorithms; Mangroves; Optical; Optical imaging; Remote Sensing; SAR; Satellites; Sea measurements; Sociology; Time series analysis; Tropical cyclones
• ISSN: 2153-7003
• DOI: 10.1109/IGARSS52108.2023.10283459

The Philippines' mangrove forests offer numerous ecosystem goods, services, and protection to coastal populations. However, in 2018, illegal and massive mangrove cutting within the areas of Bulacan was reported that caused mangrove extent changes across the province. In 2020, an initiative to plant mangrove seedlings over 76 hectares in Bulacan was taken to address perennial flooding in the villages affected by deforestation. Although mangrove cover maps for the country already exist, analysis of mangrove cover changes was limited by the challenge of acquiring cloud-free optical satellite data. This study aims to assess the spatial distribution of mangrove forests in the province of Bulacan using the combination of different bands of Landsat-8 and Sentinel-1 images arranged in time series from 2016 to 2020. A machine learning algorithm will be used to classify the images into two classes - mangroves and non-mangroves. The resultant mangrove forest maps will be used to assist the decision-making processes for rehabilitation and conservation efforts currently needed to protect and restore the mangrove forests in Bulacan province.