INDICATORS TO MONITOR FOREST DEGRADATION AND LOGGING IMPACTS IN THE BRAZILIAN AMAZON Session: 28-Mapping and monitoring tropical forest degradation with remote sensing Isabelle TRITSCH1, Lilian BLANC1, Valery GOND1, Clément BOURGOIN1, Guillaume CORNU1, Plinio SIST1 1 Institution
: CIRAD - UR Forests and Societies, 34398 Montpellier - France
Contact:
[email protected]
Introduction Different remote sensing methods to quantify forest degradation However, few spatial and temporal analysis to characterize trajectories of forest degradation through time Spatial and functional structure
Preserved forest Selective logging with RIL
Wood production forest
DEGRADATION THRESHOLD Conventional logging
Degraded forest Fire
Secondary forest Bare Soil
DEFORESTATION THRESHOLD Time
(Bourgoin, 2015)
Objectives Main objective : to propose operational indicators to monitor trajectories of forest degradation Decision tool to support forestry managers and policy makers.
Mapping the impacts of forest degradation in term of canopy opening: • logging tracks and log landings • logging gaps
Construction of spatial and temporal indicators of degradation
Analysis of the different degradation trajectories
© V. Gond
© V. Gond
Materials and methods
Study site: Municipality of Paragominas – PA (Brazil) Forestry company Cikel (140 000 ha certified by FSC since 2001). 23 logging plots in Cikel
Multi-temporal Landsat images (1991-2009) Spatial resolution of 30m – 6S corrections – max 10% cloud cover
Spectral un-mixed Landsat image: logging impacts appear in blue
Materials and methods – Remote sensing analysis
NDVI, GR Index
Bare Soil Logging tracks and log landings CLASLite (Asner et al., 2005)
Fraction Index Landsat image
(Souza et al., 2013)
Step3. Buffer 1 km
(Bourbier et al., 2013)
Senescent vegetation
to avoid confusion with natural gaps
Step1. Identification of logging tracks and log landings
Final classification
Step 4. Final classification of the canopy opening
Spectral un-mixing (S, PV, NPV)
Step2. Identification of logging gaps Step 5. Calcul of degradation metrics
Materials and methods – Remote sensing analysis
Mapping logging impacts in a forestry company (Cikel)
Year 1
Logging plots year 1 Logging plots year 2
Cloud Bare soil (logging road) Agriculture areas Logging gaps Preserved forests
Materials and methods – Remote sensing analysis
Mapping logging impacts in a forestry company (Cikel)
Year 2
Harvest of the plots of year 2
Decrease of the canopy opening in the plots of year 1
Materials and methods – Remote sensing analysis
Trajectories: important to consider time series: canopy closure is fast !
Example in a forestry company
Year 1
Year 5
Year 6
% of canopy opening
Year 4
Year 3
Year 2
Year 7
Logging plots of year 2 Logging plots of year 3 Year 0 Year 1
Year 2 Year 3
Year 4 Year 5 Year 6
Evolution of the % of canopy opening through time
Materials and methods – Remote sensing analysis Different pattern of forest degradation
Year 1
Year 4
Example in illegal logging areas
Year 2
Year 3
Year 5
Year 6
Different illegal activities: Timber harvest Charcoal Fire
30 25
Year 7
Year 8
% of canopy opening
20 15 10 5 0 Year 1
Year 2
Year 3
Year 4
Year 5
Year 6
Year 7
Year 8
Evolution of the % of canopy opening through time
Materials and methods – Indicators of forest degradation Indicators: to synthesize the evolution of the impacts through time
Frequency of the impact
Magnitude of the impact
% of canopy opening
Forestry company
- Magnitude Max = 6%
30
Max = 25%
Illegal logging
25
• • •
Maximum of canopy opening Mean % of canopy opening Cumulative % of canopy opening
- Frequency
Time
% of canopy opening
Selected indicators
20 15 10 5 0 Year 1 Year 2 Year 3 Year 4 Year 5 Year 6 Year 7 Year 8
Time
Evolution of the % of canopy opening through time
•
Number of harvest (pics)
during the whole considered period
Results
Degradation indicators in Cikel vs. in illegal logging plots during a 15 years period 2.
These simple indicators can clearly differentiate the legal logging from the illegal logging
Results
Indicators in plots certified by the FSC vs. non certified. 1. Maximum of canopy opening
2. Mean % of canopy opening
Strong impacts of FSC certification in the Cikel forestry company
3. Cumulative % of canopy opening
4. Number of harvest (pics)
Only one harvest as logging cycle is 30 years
Results
Upscalling: 73 random plots during 15 years classified in function of these indicators Four main trajectories of forest degradation
(n=73)
% of canopy opening
Cluster analysis
Time
Intensely logged plots (magnitude) (with at least one opening of ~ 20%)
Frequently degraded plots (Cumulated opening very high~ 48%)
Intermediar plots Preserved plots
Conclusion Indicators can identify different trajectories of degradation through time
Replicable methodology (arquive Landsat & Sentinel 2)
Perspectives: upscalling in larger areas
Important decision tool for forest monitoring (forestry managers, FSC)
© V. Gond
Thanks for your attention !
