Near Real-Time Change Detection System Using Sentinel-2 and Machine Learning: A Test for Mexican and Colombian Forests

Ana María Pacheco-Pascagaza,Yaqing Gou,Marco Iñiguez,Ciaran Robb,Valentin Louis,Miguel Angel Castillo Santiago,Caroline Upton,Gustavo Galindo,John F Roberts,Indira Paola Pachón Cendales,Carmen Meneses,Polyanna Da Conceição Bispo,Edersson Cabrera,Fernando D B Espírito-Santo,Heiko Balzter,Oswaldo Carrillo Negrete,Pedro Rodríguez-Veiga

doi:10.3390/rs14030707

Abstract

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-resolution 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 continents.

Highlights

Tropical forests stabilize the world’s climate and protect biodiversity [1,2]
We introduce a Sentinel-2-based near real-time (NRT) forest monitoring system that is based on an open-source Python library called Python for Earth Observation (PYEO v3.6) [43], which automates data downloading, processing, analysis, validation and change alert reporting and won the Copernicus Masters’ Award (Sustainable Living) in
In the Manantlán window area, deforestation events mainly happened in the dry season between December and March (Figure 3, Mexico)

Summary

Introduction

Tropical forests stabilize the world’s climate and protect biodiversity [1,2]. In the context of the UN Framework Convention on Climate Change and the Paris Agreement, the initiative on “Reducing Emissions from Deforestation and forest Degradation” (REDD+)aims to protect carbon stores, biodiversity and other ecosystem services [3,4]. Tropical forests stabilize the world’s climate and protect biodiversity [1,2]. In the context of the UN Framework Convention on Climate Change and the Paris Agreement, the initiative on “Reducing Emissions from Deforestation and forest Degradation” (REDD+). Aims to protect carbon stores, biodiversity and other ecosystem services [3,4]. One of the main requirements for the implementation of REDD+ is a robust and objective measurement, reporting and verification (MRV) system that is based on satellite technology [5]. Humans have changed the natural environment on a global scale to an unprecedented degree. The global changes in atmospheric chemistry and climate have led to a new geological era called the Anthropocene [6]. Land use change is the driver with the largest global impact on forest cover since three quarters of these lands have changed, thereby diminishing the productivity of 23% of global land surface [4,7,8]

Methods

Results

Discussion

Conclusion