Infoterra presented at IMAZON** side-event at COP15 « REDD Making it happen: Monitoring forest emissions and governance to achieve REDD » (more) a demonstration of large-scale monitoring of deforestation and degradation in the Amazon basin, sponsored by EADS Astrium.
The uniqueness of this initiative stands in developing an image processing suite, with methodology backed by scientific community, that enables continuous & operational monitoring of forest changes over time.
Outcomes of the project demonstrate the use of satellite images to establish the reference situation of Forest in a given country (the REDD Baseline) and to develop operational concepts for Monitoring and Verification.
The concrete achievement encompasses:
- a quantitative and dynamic analysis of deforestation & forest degradation over 25 years in Mato Grosso (forest biome covering more than 600,000 km² in this state)
- an accurate carbon emission assessment associated to deforestation & degradation
A unique technology was developed to transform images into value-added information and address the increasing demand for an industrial solution that can be deployed rapidly on any forest basins. Built upon Overland, this paves the way for:
- the development of permanent standards to monitor status and trends of rainforests worldwide
- the sustainable management of forest carbon stock
- the transfer of space & forestry technology & know-how for country accountability on forest carbon stock and its value on financial markets, etc.
Such tools are essential to support the financial mechanism that will allow to compensate actions on forest conservation.
The remote sensing solutions being developed for REDD are based on methodology using reference Forest indicators (e.g. NDFI***) and providing reproducible results from multiple sensors.
BASELINE: Processing large volumes of image archives to capture the Forest dynamics over time through annual land cover maps
- Landsat archive collected and pre-processed by IMAZON over the whole Amazon basin (25 years).
- Demonstration on Mato Grosso state in Brazil : production of the historical situation (mid 80's) and trend since this date : 650 Landsat images processed in less than 2 months. Production was performed jointly between Infoterra and IIMAZON
- Focus on the comprehensive mapping of rainforest / deforested / degraded/ forest regeneration areas, and of their further evolution
Figure 1: Methodology - definition of target nomenclature
MONITORING & VERIFICATION: Operational solutions for Monitoring and Verification can be developed thanks to:
- An access to a large range of space sensors with various swaths, resolutions and spectral bands
- The use of High Resolution (HR) to automatically identify degradation (detect even minor logging events) and to improve accuracy of Forest Land Cover (LC) mapping
- The use of Very High Resolution (VHR) data to perform routine quality control and verification, through ultimate photo-interpretation
- The use of wide-swath sensors to perform global monitoring and near real-time detection of newly deforested areas
Figure 2: Monitoring with High Resolution images (SPOT5) - increased logging detection using the NDFI map (click here to unlarge the image)
Figure 3: Validation of Forest LC map with VHR true colour SPOT5 image - 2,5m (click here to unlarge the image)
Figure 4: Near-Real Time monitoring with wide-swath sensor (MERIS) - Producing the NDFI maps with MERIS and comparison with Landsat (click here to unlarge the image)
*REDD Reducing Emission from Deforestation & Degradation
**The Amazon Institute of People and the Environment
***Normalized Difference Fraction Index, Forest indicator developed by IMAZON (C. Souza)
