A Pantropical Near Real Time Monitoring
System for Vegetation Change
METHODOLOGY
WHAT IS THE TERRA-i TOOL?
•A decision support system in operation for land use and
land cover information based on earth observation ‘big data’
analytics.
•Detects vegetation loss resulting from human activities.
•Works in near real-time, producing maps every 16 days.
Minimum change area detected: 6.25 ha
Temporal coverage: from 2004 to present
•Goal: inform government institutions, conservation
organizations and the general public about where and when
natural vegetation loss occurs.
•Premise: natural vegetation loss follows a predictable
pattern of change in greenness from one date to the next,
brought about by site specific characteristics and climatic
conditions over the preceding days.
•Bayesian-probability based neural network (BNN) learns
how the greenness of a given pixel responds to a unit of
rainfall (year 2000-2004).
•Calibrated model is run to identify fluctuations in greenness
that cannot be explained by rainfall or by previous state of the
vegetation (year 2004-present).
•Anomalies are flagged as potential vegetation loss.
RESULTS - MOVING TOWARDS DATA GENERATION
Identifying hotspots of change
Beyond humid tropical ecosytems
Modeling future scenarios of change
Terra-i can pinpoint areas that are experiencing
rapid landscape changes. Cash crops, shifting
agriculture and mining are the primary drivers of
local deforestation in the Western Amazon.
Other non-humid habitats have suffered losses.
Terra-i data has revealed deforestation in the
Dry Chaco that is much more severe than in the
Amazon.
Maps of potential deforestation or other natural
vegetation loss due to new infrastructure
construction can be obtained based on Terra-i
data. A pilot model has been tested in Guyana.
Achievements
In the 10 years since Terra-i’s inception, ~260 organisations from 45 countries have
reported using its data as part of research, conservation and development efforts.
Terra-i is one of the principal sources of natural vegetation monitoring information for
the Global Forest Watch initiative and the Ministry of Environment of Peru.
A new crowdsourcing game, ‘Forest Defenders’, proved to be a useful way to validate
Terra-i remote sensing methodology.
Terra-i innovation and relevance has been recognised with the GeoSUR award, a prize
given by the Pan-American Institute of Geography and History (IPGH).
APPLICATIONS
•Monitoring effectiveness of conservation.
•Estimating environmental impacts of infrastructure.
•Assessing contribution of deforestation to climate change.
•Understanding impact of recent land cover change on
ecosystem services (through WaterWorld and Co$ting Nature
Policy Support Systems).
Terra-i is a collaboration between multiple international organisations led by the
International Center for Tropical Agriculture - CIAT. Currently, Terra-i is funded by the
CGIAR Research Program on Forests, Trees and Agroforestry and GFW initiative.
Contact point at King’s College London: Mark Mulligan, [email protected]
FUTURE PROSPECTS
•Examining post-deforestation land use trajectories for land
use change impacts assessment
•Self-monitoring of deforestation and other natural vegetation
loss by lower-income countries.
•Ensuring the tool’s sustainability through community-based
analysis of results.