integrating emission factor and activity data assessment ......emission levels (rel) and the upper...
TRANSCRIPT
Integrating Emission Factor and Activity
Data Assessment in Support of REDD+
MRV
Ben DeVries
7 May, 2013
What is REDD+?
RED
Reducing Emissions from Deforestation in developing
countries
UNFCCC – determines the “rules”
IPCC – Good Practice Guidelines (GPG)
for reporting in various sectors,
including Land Use, Land Use Change,
and Forestry (LULUCF)
REDD
Reducing Emissions from Deforestation and forest
Degradation in developing countries
REDD+
Reducing Emissions from Deforestation and forest
Degradation and the role of conservation, sustainable
management of forests and enhancement of forest carbon stocks in developing countries
Measuring, Reporting and Verification (MRV)
C Stock Change (t/ha)
Emission
Factor
(EF)
Area Change (ha)
Emissions
(E)
Activity
Data
(AD)
Car
bo
n s
tock
Baseline (reference)
Time
Reduced Emissions
Reported
Activity data: land area
affected by change (e.g. area
of forest cleared)
Activity monitoring system is
important
Emission Factor: amount of carbon
released per unit area as a result of
activity
Site-specific quantification of carbon
stocks, or use of IPCC default factors
Some Considerations for REDD+ MRV
Participating countries will need to develop a working definition of “forest” (ie. height, canopy density, and area thresholds)
Reference Emission Levels (aka Reference Levels; REL or RL) need to be determined:
● Based on historical emissions due to deforestation and degradation, or on modeled projected emissions into the future
● REDD+ interventions should demonstrate additionality: are emissions relative to RL’s actually due to interventions? Or other phenomena?
Conservativeness is a contentious point in REDD+ discussions.....
A Conservative Approach to MRV
Grassi G., et al (2008). Environmental Research Letters, 3.
A conservative approach to estimating emissions reductions reduces the risks of crediting false
emission reductions.
Conservative estimate: The difference between the lower confidence bound of Reference
Emission Levels (REL) and the upper confidence bound of the reported emissions in the
assessment period.
Consequences of High Uncertainty
Simulated uncertainties based
on REL data
Only one projected scenario
results in emissions reductions
beyond the confidence interval
(after 2020)
Pelletier, J., Ramankutty, N. & Potvin, C. (2011). Environmental Research Letters, 6
High uncertainties are
often a result of lack of
data (b/c of lack of funds,
capacities, etc.)
This could undermine the
implementation of REDD+!
How can you prove that
emission reductions are
real [i.e. additionality]?
The Local Context
Ethiopia is embarking on a national REDD+ process, which includes building up a national MRV and NFM system
There are alot of capacity gaps to be addressed at the national level it will be important to scale up [existing] local monitoring activities
National REDD+ MRV System
Local (e.g. project-based) forest monitoring activities / systems
Methods, data,
experiences
The Local Context
NABU Forest and Climate Monitoring Project
Funded through the International Climate Initiative (ICI) by the German Federal
Ministry for the Environment, Nature Conservation and Nuclear Safety (BMU)
This research falls under a broader implementation project carried out by NABU
Germany / Ethiopia and local partners in Kafa, Southwest Ethiopia:
Climate Change Mitigation and Primary Forest Conservation – A Best-Practice
Management Scheme for Wild Coffee Forests in Ethiopia
General project objective is to monitor the impact of REDD+ related implementation
activities on carbon emissions in the project area
UNESCO Kafa Coffee Biosphere Reserve http://www.kafa-biosphere.com/ http://www.nabu.de/en/aktionenundprojekte/kafa/
Activity Data
RS Time series data in situ monitoring data
Change
Maps
Reported
Disturbances
Emission Factor
LC maps
Stratified
sampling
Measured
Biomass
Towards an Integrated Monitoring System
C STOCK CHANGES
Monitoring Forest Change from Space
Bitemporal change detection
(MAD; 2000-2010)
Time-series analysis (BFM;
2005-2011 annual)
Spatiotemporal Activity Monitoring
Breakpoints defined as a statistical
deviation from a stable historical time
series
Timing (x-axis) of the breakpoint is
an important output of the method
Change magnitude
defined as the median
of the residuals from
the observed and
expected time series
BFAST Monitor:
http://cran.r-project.org/web/packages/bfast/bfast.pdf
Are these parameters
related to
deforestation/degradation?
TS Workflow
Assessing the Dynamics of Change
BFM run using successive monitoring
windows (1-year monitoring periods)
magn (median residual of NDVI
magnitudes between expected and
observed values in monitoring period)
Hypothesis: magn parameter is
related to intensity of change, and
could be used to detect degradation
(to a limited extent)
How does choice of metric affect
sensitivity of the method?
Carbon Stocks (and Stock Changes)
CO2 CO2 CO2
CO2 Carbon Stocks and Flows
Biomass Mapping (Pan-Tropical)
Baccini, A., N. Laporte, S. Goetz, M. Sun, W. Walker, J. Kellndorfer, R.A. Houghton (2009). Pantropical Forest Carbon Mapped with Satellite and Field Observations. The Woods Hole Research Center, Falmouth, MA, USA. A. Baccini, S J. Goetz, W. Walker, N. T. Laporte, M. Sun, D. Sulla-Menashe, J. Hackler, P. Beck, J. Kellndorfer, M. Friedl, R. A. Houghton. Submitted (2010)
Very coarse resolution (500m)
How to capture local variability in carbon
stocks? How to assess C-stock changes?
Tier 1
Targeted Stratification to Assess EF
Tier 3 Objectives:
1) Improve existing (tier 1) biomass products
2) Specifically target change areas (def/deg) and
areas expected to undergo change (risk)
‘risk’ is defined based on the average
distance within which 75% of change
occurred in subsequent years (i.e.
where is change in 2013 expected to
occur?)
Stratified Random Sampling
Strata include already
deforested or degraded areas
[obj 1] as well as expected
deforested (“at-risk”) areas
[obj 2].
A B
C ADref
E = AD X EF Model scenarios: 1) actual change only happens in stratum C (AD ϵ C) use only EFC and compare with classical scenario (where EFA is used):
E1 = AD X EFA | E2 = AD X EFC
2) actual change happens partly in stratum C and partly in stratum B (“unexpected” change) use EFC and EFB in the calculation and compare with classical scenario (EFA)
E1 = AD X EFA | E2 = (ADC X EFC) + (ADB X EFB)
1. Classical 2. Targeted EFA EFB
EFC
Impact of Targeted Stratification
Summary
High uncertainties in carbon stocks and carbon stock changes are a major concern for REDD+ MRV
Integration of data streams used to assess Activity Data and Emission Factors can help to improve emissions estimates where C-stock data are absent or very difficult to obtain
Using Activity Data estimates derived from high temporal-resolution change maps (e.g. annual and near real-time estimates), forest areas could be stratified based on the expected areas of change
Further Reading
Ethiopia CRGE:
● http://www.epa.gov.et/Download/Climate/Ethiopia's%20Climate-Resilient%20Green%20economy%20strategy.pdf
Ethiopian R-PP:
● http://www.forestcarbonpartnership.org/sites/forestcarbonpartnership.org/files/Documents/PDF/Sep2010/Ethiopia_draft_R-PP_August_2010.pdf
Herold, M., Roman-Cuesta, R.M., Mollicone, D., Hirata, Y., Laake, P. Van, Asner, G.P., Souza, C., Skutsch, M., Avitabile, V. & Macdicken, K. (2011) Options for monitoring and estimating historical carbon emissions from forest degradation in the context of REDD+. Carbon balance and management, 6, 13.
Pelletier, J., Ramankutty, N. & Potvin, C. (2011) Diagnosing the uncertainty and detectability of emission reductions for REDD + under current capabilities: an example for Panama. Environmental Research Letters, 6.
Verbesselt, J., Zeileis, A., & Herold, M. (2012). Near real-time disturbance detection using satellite image time series. Remote Sensing of Environment, 123, 98–108.
