TY - JOUR
T1 - Near Real-Time Change Detection System Using Sentinel-2 and Machine Learning: A Test for Mexican and Colombian Forests
AU - Pascagaza, Ana Maria Pacheco
AU - Gou, Yaqing
AU - Louis, Valentin
AU - Roberts, John F.
AU - Rodriguez-Veiga, Pedro
AU - Bispo, Polyanna da Conceição
AU - Espírito-Santo, Fernando
AU - Robb, Ciaran
AU - Upton, Caroline
AU - Galindo, Gustavo
AU - Cabrera, Edersson
AU - Cendales, Indira Paola Pachón
AU - Santiago, Miguel Angel Castillo
AU - Negrete, Oswaldo Carrillo
AU - Meneses, Carmen
AU - Iñiguez, Marco
AU - Balzter, Heiko
N1 - Funding Information:
Acknowledgments: We acknowledge to the Global Challenges Research Fund and the UK Space Agency’s International Partnership Program to fund the Forests 2020 project. Also supported by EASOS and the National Centre for Earth Observation (NCEO). Forest Sentinel was supported by NERC “REDD+ Monitoring Services with Satellite Earth Observation” (NE/N017021/1). All data were processed on the ALICE high performance-computing cluster managed by the University of Leicester. We thank all partners within the projects for their collaboration, data sharing and infrastructure. We also thank the European Commission Copernicus and Planet Team for the free and open data.
Funding Information:
This research was part of the Forest 2020 project funded by the Global Challenges Research Fund for the UK Space Agency?s International Partnership Program, Forest 2020 project, within the frameworks of the Earth and Sea Observation System (EASOS) Malaysia project and the Forests 2020 project. This work was also supported by the Natural Environment Research Council?s National Centre for Earth Observation (NCEO). Acknowledgments: We acknowledge to the Global Challenges Research Fund and the UK Space Agency?s International Partnership Program to fund the Forests 2020 project. Also supported by EASOS and the National Centre for Earth Observation (NCEO). Forest Sentinel was supported by NERC ?REDD+ Monitoring Services with Satellite Earth Observation? (NE/N017021/1). All data were processed on the ALICE high performance-computing cluster managed by the University of Leicester. We thank all partners within the projects for their collaboration, data sharing and infra-structure. We also thank the European Commission Copernicus and Planet Team for the free and open data.
Publisher Copyright:
© 2022 by the authors. Licensee MDPI, Basel, Switzerland.
PY - 2022/2/2
Y1 - 2022/2/2
N2 - The commitment by over 100 governments covering over 90% of the world’s forests at the COP26 in Glasgow to end deforestation by 2030 requires more effective forest monitoring systems. The near real-time (NRT) change detection of forest cover loss enables forest landowners, government agencies and local communities to monitor natural and anthropogenic disturbances in a much timelier fashion than the thematic maps that are released every year. NRT deforestation alerts enable the establishment of more up-to-date forest inventories and rapid responses to unlicensed logging. The Copernicus Sentinel-2 satellites provide operational Earth observation (EO) data from multi-spectral optical/near-infrared wavelengths every five days at a global scale and at 10 m resolution. The amount of acquired data requires cloud computing or high-performance computing for ongoing monitoring systems and an automated system for processing, analyzing and delivering the information promptly. Here, we present a Sentinel-2-based NRT change detection system, assess its performance over two study sites, Manantlán in Mexico and Cartagena del Chairá in Colombia, and evaluate the forest changes that occurred in 2018. An independent validation with very high-reso-lution PlanetScope (~3 m) and RapidEye (~5 m) data suggests that the proposed NRT change detection system can accurately detect forest cover loss (> 87%), other vegetation loss (> 76%) and other vegetation gain (> 71%). Furthermore, the proposed NRT change detection system is designed to be attuned using in situ data. Therefore, it is scalable to larger regions, entire countries and even con-tinents.
AB - The commitment by over 100 governments covering over 90% of the world’s forests at the COP26 in Glasgow to end deforestation by 2030 requires more effective forest monitoring systems. The near real-time (NRT) change detection of forest cover loss enables forest landowners, government agencies and local communities to monitor natural and anthropogenic disturbances in a much timelier fashion than the thematic maps that are released every year. NRT deforestation alerts enable the establishment of more up-to-date forest inventories and rapid responses to unlicensed logging. The Copernicus Sentinel-2 satellites provide operational Earth observation (EO) data from multi-spectral optical/near-infrared wavelengths every five days at a global scale and at 10 m resolution. The amount of acquired data requires cloud computing or high-performance computing for ongoing monitoring systems and an automated system for processing, analyzing and delivering the information promptly. Here, we present a Sentinel-2-based NRT change detection system, assess its performance over two study sites, Manantlán in Mexico and Cartagena del Chairá in Colombia, and evaluate the forest changes that occurred in 2018. An independent validation with very high-reso-lution PlanetScope (~3 m) and RapidEye (~5 m) data suggests that the proposed NRT change detection system can accurately detect forest cover loss (> 87%), other vegetation loss (> 76%) and other vegetation gain (> 71%). Furthermore, the proposed NRT change detection system is designed to be attuned using in situ data. Therefore, it is scalable to larger regions, entire countries and even con-tinents.
KW - Deforestation
KW - Machine learning
KW - Near real-time
KW - Tropical forests
KW - Vegetation change detection
UR - https://www.mdpi.com/2072-4292/14/3/707
U2 - 10.3390/rs14030707
DO - 10.3390/rs14030707
M3 - Article
SN - 2072-4292
VL - 14
JO - Remote Sensing
JF - Remote Sensing
IS - 3
M1 - 707
ER -
