• Home
  • What stories can a pixel tell?: monitoring landscape changes through remote sensing

What stories can a pixel tell?: monitoring landscape changes through remote sensing

A bird’s-eye view shows patterns and changes not so easily apparent from the ground. Photo: World Agroforestry Centre/Yudi Noviandi
Posted by


A bird’s-eye view shows patterns and changes not so easily apparent from the ground. Photo: World Agroforestry Centre/Yudi Noviandi

By Amy Cruz, originally published at ICRAF’s Agroforestry World Blog

A monitoring system has the potential to help governments, research organizations and other groups detect changes in landscapes, such as deforestation and establishment of oil palm plantations.

Representatives from government agencies, research organizations and academe in the Philippines attended a forum on the near real-time monitoring of forest landscapes through the Terra-i Remote Sensing System. The event provided the participants a platform for discussing potential uses of the Terra-i system in the country.

Two sessions were jointly organized by the International Center for Tropical Agriculture (CIAT) and ICRAF The World Agroforestry Centre at the Diliman and Los Baños campuses of the University of the Philippines in October 2016. Louis Reymondin of CIAT explained the system, after which the participants discussed potential collaboration and uses in the country.

Robustly used in Latin America, the Terra-i system has recently gained interest in Southeast Asian countries such as Cambodia, Myanmar, the Philippines and Viet Nam as a near real-time monitoring system for habitat loss and landscape changes. The goal of the system is to detect as soon as possible changes in a landscape but not to exactly measure forest loss in hectares because of the limitations of the satellite sensor.

Terra-i leverages remote sensing and advanced machine-learning techniques to detect changes in a landscape. According to Reymondin, a team of researchers developed the system based on the premise that natural vegetation follows a predictable pattern of changes in greenness over time. These changes are based on site-specific characteristics and the climatic conditions of preceding days.

Read also: Terra-i, new data – new look for the website

The system predicts how green a pixel should be, given a unit of rainfall, and compares that specific pixel with satellite images. Any anomalies in the time series, which could be due to human activities, can be identified. The time series are produced in 250 m spatial resolution every 16 days to reduce the effects of cloud cover. Although it sounds complicated, Terra-i was designed to be implemented even with limited hardware and financial resources.

Such a system would help improve forest management, reduce emissions from deforestation and forest degradation and contribute to climate-change adaptation and mitigation. Some of the possible applications include detecting mining activities and new oil-palm plantations; assessing the effectiveness of protected areas; and developing potential deforestation scenarios from trends.

Other activities in the pipeline for Terra-i include:

  1. development of 30 m-resolution time series through combining pictures from different satellite systems;
  2. identification and characterization of degradation through combining images from satellites and unmanned aerial vehicles; and
  3. analyses of time series to tell ‘the story of the pixel’, for example, how are the changes spatially distributed and what happens after deforestation.

The team hope that later the system can be modified to monitor other types of ecosystems in the tropics. To achieve this, they will need local knowledge and data to calibrate pan-tropical data available from open-access satellite imagery.

This work is linked to the CGIAR Research Program on Forests, Trees and Agroforestry.

  • Home
  • Terra-i team, scientific talent for a greener world

Terra-i team, scientific talent for a greener world

Posted by


Terra-i_EN_300-1Originally published at CIAT’s blog

The Terra-i,  which works under CIAT’s Decision and Policy Analysis (DAPA) Research Area, is constantly evolving, and their main goal this year is to process a larger amount of data in less time.

“To meet this challenge we purchased more data processing devices and improved the software of our current equipment to increase automatization of some of the steps in the process. He also added that this work was carried out in joint collaboration with the College of Engineering and Management of the Canton of Vaud (HEIG-VD, based in Switzerland).” Louis Reymondin, Leader of the Terra-i team

Terra-i is a vegetation monitoring system for Latin America, Africa, Asia and Oceania, used to detect land-cover changes in near real time. This tool is based on the hypothesis that vegetation has predictable patterns with specific cycles according to climatic conditions and soil characteristics.

The Terra-i system generates specialized information for sustainable use and conservation of natural resources and ecosystems. Moreover, it is updated permanently and allows to incorporate the results from analyses with higher resolution images. This information is used to address climate change mitigation, forest conservation, and biological diversity, among others.

It is important to highlight that this initiative is an inter-institutional collaboration betweenCIAT, The Nature Conservancy (TNC), the CGIAR Research Program on Forests, Trees and Agroforestry (FTA), the College of Engineering and Management Vaud (HEIG-VD, based in Switzerland), and King’s College London (KCL, based in the UK).

“This year we are planning to obtain more data for analysis in less time, particularly in Latin America and the tropics. We are searching for new opportunities in countries such as Honduras and Guatemala, and we will present the system in Asia to form a new working team.” Paula Andrea Paz, Research Assistant

Terra-i’s evolution

Recent achievements

  • The Peruvian Ministry of Environment (MINAM) adopted the Terra-i system to monitor deforestation.
  • A group of American scientists used the data collected between 2004 and 2012 in eastern Honduras to write a publication in Science.
  • Data shared with the “cartochaco” initiative. For further information, please visit www.cartochaco.org.
  • The team became a data provider for the Global Forest Watch, one of the biggest initiatives in data analysis on forest loss.
  • Moreover, Global Forest Watch allocated resources to the Terra-i team to run analyses and produce data every month instead of every two months.
  • The team received the CIAT Award for Outstanding Research Outcome, with a cash price of US$5.000 to fund their professional development.

Our team: 

Louis Reymondin

Louis Reymondin, Leader of the Terra-i team

Louis Reymondin is an expert in the development of software that combines big data and georeferenced information. He studied software development in the University of Applied Sciences in West Switzerland and then completed a PhD in Geography in the King’s College London. Louis has more than 10 years of experience in the coordination of the Terra-i project and research that involves the use of artificial intelligence and data mining to better understand the dynamics between human and environmental activities.

Jhon Jairo Tello

Jhon Jairo Tello, System Engineer

Jhon Jairo Tello graduated from Universidad Cooperativa de Colombia since 2011. Since then, he has been working on Terra-i, mainly in the development of its website.

Paula Andrea Paz

Paula Andrea Paz, Research Assistant

Topographical Engineer from Universidad del Valle. She works on the Terra-i project with data and information download, processing and post-processing, and supports deforestation monitoring in Latin America and across the tropics.

Oscar Bautista

Oscar Bautista, Research Assistant

Agricultural Engineer from Universidad Nacional de Colombia. He has been working since 2012 as a consultant in diverse conservation projects. Moreover, he has been engaged in the identification of ecosystem pressure and threats to assess land-use change and forest loss impact. Currently, he works as a research assistant in DAPA and is part of the team responsible for data generation, download, and processing.

Bernadette Menzinger

Bernadette Menzinger, Visiting Researcher

Geographer graduated from Julius-Maximilians-Universitat Wurzburg in Germany in 2014. She specialized as a research assistant supervising land use in West Africa. Currently she is pursuing her masters studies in High Mountain Geography and Climatology at the University of Graz in Austria, and is working as a visiting researcher in Terra-i. She is responsible for updating data, and downloading and processing information.

Finally, it is important to highlight the work carried out by former members of the Terra-i team who contributed to the success of this project. The Terra-i team wants to thank especially Karolina Argote and Alejandro Coca for their great contributions and dedication to the project.

More info about the Terra-i team
Terra-i: A Cool Tool for Detecting Deforestation. 

  • Home
  • Measure it and manage it: Terra-i forest monitoring goes global

Measure it and manage it: Terra-i forest monitoring goes global

Members of the Terra-i team discuss the Terra-i deforestation monitoring system, which can zoom-in on Latin America's forest to track deforestation in near real-time. Photo: Neil Palmer/CIAT
Posted by


By Ruben Echeverria, Director General, Center International Center for Tropical Agriculture (CIAT)

Members of the Terra-i team discuss the Terra-i deforestation monitoring system, which can zoom-in on Latin America's forest to track deforestation in near real-time. Photo: Neil Palmer/CIAT
Members of the Terra-i team discuss the Terra-i deforestation monitoring system, which can zoom-in on Latin America’s forest to track deforestation in near real-time. Photo: Neil Palmer/CIAT

The Amazon rainforest is often described as the ‘lungs of the Earth’. But the forests of South East Asia and Africa are also vitally important – and among the most at risk. That’s why a deforestation early-warning system that’s proving so successful in Latin America will soon be monitoring all the world’s tropical forests – from space. By combining detailed satellite images with a lot of number crunching, the deforestation monitoring system Terra-i harnesses the power of big data to help protect forests, biodiversity, ecosystem services and livelihoods.

With images updated and scrutinized every 16 days, it can distinguish between recent and historical deforestation back to 2004, giving a frighteningly accurate story of forest clearance. It also enables users to identify the drivers of forest loss – from agriculture to mining, road building, urbanization and more.

Launched in Latin America in 2012, the Government of Peru adopted Terra-i as its official deforestation monitoring system two years later. Since then it has been keeping watch over its share of the Amazon rainforest, and flagging new drivers in 2015. Several countries in Central America are poised to adopt it.

Last year also saw a lot of hard work to prepare for the launch of Terra-i in SE Asia and Africa – a move that means the system will soon be watching over all the world’s tropical forests. In SE Asia, we expect early adopters to be Vietnam, Laos and Cambodia – where deforestation rates are among the highest in the world.

The system will also track forest clearance and vegetation change Africa, specifically in the enormously rich, diverse and important Congo Basin.

While some of the images generated by the system can be unnerving, the team behind Terra-I know that when it comes to deforestation, if you can’t measure it you can’t manage it. By giving a clearer idea of where the hotspots are, and what’s likely to be causing them, Terra-i enables governments to develop more robust policies on forest protection.

It can also help them quantify the enormous carbon dioxide emissions generated by forest clearance, meaning they can put a more accurate price on conservation.

The private sector can benefit from Terra-i too, with businesses better able to assess the environmental impact of their activities, which in turn can feed into their corporate social responsibility programs.

These activities go to the heart of the important issues that the CGIAR Research Program on Forests, Trees and Agroforestry is committed to tackling.

The expansion of Terra-i to SE Asia and Africa represents a big step towards uniting scientists and policymakers across the tropics behind a common goal: the protection and management one of our most precious and vulnerable resources.

Terra-i is the result of collaboration between CIAT, Kings College London and the University of Applied Sciences of Western Switzerland (HEIG-VD). It is funded by The Nature Conservancy, The World Resources Institute, Global Forest Watch and the CGIAR Research Program on FTA.


Back to top

Sign up to our monthly newsletter

Connect with us