Capturing the impacts of human activities on reported forest greenhouse gas emissions Werner A. Kurz Natural Resources Canada Canadian Forest Service 2025 and Beyond: Modeling the land sector’s role in achieving GHG mitigation and adaptation goals 8 th Forestry and Agriculture Greenhouse Gas Modeling Forum Shepherdstown, Sept 26 - 28, 2016
41
Embed
Capturing the impacts of human activities on reported forest greenhouse gas emissions · 2017-09-11 · Capturing the impacts of human activities on reported forest greenhouse gas
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Capturing the impacts of human activities on reported forest greenhouse gas emissions
Werner A. Kurz
Natural Resources CanadaCanadian Forest Service
2025 and Beyond: Modeling the land sector’s role in achieving GHG mitigation and adaptation goals
8th Forestry and Agriculture Greenhouse Gas Modeling ForumShepherdstown, Sept 26 - 28, 2016
AcknowledgementsThe approach presented here is the result of the collaboration between the Canadian
Forest Service of Natural Resources Canada and the Pollutant Inventories and Reporting
Division of Environment and Climate Change Canada.
ECCC:
Doug MacDonald
Dominique Blain
Shari Hayne
Ana Blondel
NRCan:
Werner Kurz
Tony Lemprière
Max Fellows
Scott Morken
Carolyn Smyth
Mark Hafer
Eric Neilson
Karin Simonson
Alison Beatch
2
Outline
Background
Motivation
Methods
Results
Discussion
Conclusions
3
Background
Managed Land Proxy (MLP): IPCC default approach
The IPCC Managed Land Proxy assumes that emissions and
removals from anthropogenic activities in the land sector can be
approximated as those from managed lands.
However, natural disturbances contribute large direct and indirect
(delayed) emissions. These contribute to large interannual
variability and trends in reported estimates.
The impacts of natural disturbances mask trends in emissions
resulting from changes in human activities, including mitigation
efforts.
4
Motivation
Develop a method that better reflects the impacts of human
activities and integrated land management on trends in reported
greenhouse gas emission inventories in Canada’s forest sector.
Our analysis is exploring methods to exclude the emissions and
subsequent removals of greenhouse gasses resulting from
uncontrollable natural disturbances such as wildfire, insects and
storms in the managed forest (i.e. forest land remaining forest land).
5
IPCC methodological options
The 2010 IPCC report from the Expert
Meeting Report on “Revisiting the
Use of Managed Land as a Proxy for
Estimating National Anthropogenic
Emissions and Removals”:
reviewed the MLP shortcomings,
introduced five possible
approaches to estimating
anthropogenic emissions in the
land sector, and
concluded that further work is
required before any alternative to
the MLP can be recommended.
6
IPCC 2010 lists five methodological options
Maintenance of the managed land proxy
Component separation (quantify the influence of different drivers and
then identify which drivers contribute to anthropogenic emissions)
Comparison of two time series (that represent two different levels of
human activities, e.g. current management and no management, or
improved management vs. business as usual management)
Default factors and optimal fingerprinting (without quantifying the
relative contributions, apply a default factor which indicates the impact
of the human activity)
Activity based approach (estimate emissions by different activities and
sum up anthropogenic contributions)
Source: IPCC 2010
7
IPCC 2010 Conclusions
[T]he managed land proxy has several shortcomings
For some countries and circumstances, use of the managed land
proxy may lead to emission and removal estimates dominated by
natural effects occurring on managed land and this would need to
be recognised where inventory estimates were used in estimates of
anthropogenic or management effects.
[T]here is also a requirement to be able to identify the impact of
mitigation and management efforts even where these are
overwhelmed by the impacts of natural processes (e.g. natural
disturbances) or where these are obscured by inter-annual variations
in greenhouse gas fluxes.
[F]urther work by the scientific community will result in more mature
approaches which can be assessed at a later date.
8
Outline
Background
Motivation
Methods
Results
Discussion
Conclusions
9
A new approach
The 2010 IPCC Expert Meeting Report introduced the notion of
“Component Separation” but focussed on “drivers of fluxes” such as
CO2 fertilization, atmospheric N-deposition and climate change (i.e.
the issues addressed in the “factoring out” discussions.)
The report also introduced an activity-based approach in which the
emissions associated with anthropogenic activities in the land sector
are estimated and summed.
Here we develop a new combined approach that allows us to estimate
separately the emissions and removals from
1. lands subject to forest management, and
2. lands affected by natural disturbances (causing >20% mortality),
10
A new approach
Is a land-based approach.
Estimates the E/R for the entire managed forest (FL-FL).
E/R from lands affected by stand-replacing ND are excluded from
reporting for 60 years.
E/R from lands affected by ND causing >20% mortality after 1990
are excluded from reporting until pre-disturbance aboveground
biomass is reached.
Therefore, the remaining reported E/R better quantify the impacts
of forest management, including mitigation activities.
Forest management activities on excluded lands cause re-entry
into reporting.
The sum of reported and excluded emissions is equal to the
reported FL-FL values in the conventional national forest GHG
inventory.
11
Explanation of concept
Applies to all managed forest land.
12
Managed Forest Land (FL-FL)
Explanation of concept
All stands originate either from stand-replacing natural disturbance
(ND, typically fire) or clearcut harvest.
13
Managed Forest Land (FL-FL)
Harvest Origin ND Origin
(>=60 yrs old)
Explanation of concept
Stand-replacing wildfire excludes stands from reporting for 60 years
Red = excluded from reporting
Green = reported
14
Managed Forest Land (FL-FL)
Harvest OriginND Origin
(>=60 yrs old)
ND Origin
(<60 yrs old)
Explanation of concept
Natural disturbances after 1990 that cause > 20% biomass mortality
excluded from reporting until pre-disturbance biomass is reached.
Red = excluded from reporting
Green = reported
15
Managed Forest Land (FL-FL)
Harvest Origin ND Origin
(>=60 yrs old)
ND Origin
(<60 yrs old)
ND >20% mortality after 1990
(not yet at pre-disturbance biomass)
Explanation of concept
Lands affected by ND > 20% mortality after 1990 re-enter the
reporting when they reach pre-disturbance biomass.
Red = excluded from reporting
Green = reported
16
Managed Forest Land (FL-FL)
Harvest Origin ND Origin
(>=60 yrs old)
ND Origin
(<60 yrs old)
ND >20% mortality after 1990
( < pre-disturbance biomass)
ND >20% mortality
after 1990
(> pre-disturbance biomass)
Explanation of concept
ND lands affected by forest management activities (e.g. salvage
logging, planting, rehabilitation) re-enter the reporting.
Red = excluded from reporting
Green = reported
17
Managed Forest Land (FL-FL)
Harvest Origin ND Origin
(>=60 yrs old)
ND Origin
(<60 yrs old)
ND >20% mortality after 1990
( < pre-disturbance biomass)
ND >20% mortality
after 1990
(> pre-disturbance biomass)
ND followed
by FM
Decision tree for the separation of lands In the initial forest inventory all stands in FL-FL are established either
following a stand-replacing natural disturbance (ND) or following harvest.
Stand
established
following
ND?
Include in reporting
Stand
reached
pre ND
biomass
stocks?
>= 60
years since
ND?
Partial ND
with > 20%
mortality?
Exclude from reporting
Yes
Yes
No
No
Yes No
Has any
FM activity
occurred
since ND?
No
No
YesYes
18
Implementation through separation of land fluxes
The CBM-CFS3 is used to estimate emissions and removals in
Canada’s forest sector.
The model allows for the estimation of emissions and removals in about
2.7 million stands across Canada’s managed forest.
The activity data time series starts in 1990 – we do not have data on
partial disturbances prior to 1990.
The 1990 inventory identifies for each stand the age and whether or not
the stand was established from wildfire or harvest.
The goal of the implementation is to separate the emissions and
removals into reported and excluded fluxes.
To obtain a better understanding of the contribution to the reported and
excluded fluxes we stratified the stands into seven land components.
19
The seven component fluxes represent managed forest lands subject to human or natural
disturbances:
Harvest
1. Clear cut harvest before 1990 (included)
2. Clear cut or partial harvest after 1990 (included)
Stand replacing NDs
3. Before 1990, and more than 60 years ago (included)
4. Before 1990, less than 60 years old (excluded)
5. After 1990, less than 60 years old (excluded)
*. [ After 1990, more than 60 years old (would be included but is not observed) ]
Partial NDs
6. After 1990, reached pre-disturbance AGB (included)
7. After 1990, not yet reached pre-disturbance AGB (excluded)
Reported fluxes are the sum of categories 1, 2, 3 and 6
Excluded fluxes are the sum of categories 4, 5 and 7
20
Implementation through separation of land fluxes
Implementation
Technical details
During a model simulation, stands affected by natural or human
disturbances are assigned a “flag”.
Stands are monitored in each time step to determine if certain rules
(age or biomass criteria) are met and if yes “flags” are updated.
When the results database is compiled, fluxes are summarised by time
step and various reporting categories, including FL-FL, administrative
and ecological reporting zones, as well as the seven components.
21
Carbon Budget Model of the
Canadian Forest Sector (CBM-CFS3) An operational-scale model of forest C dynamics.
Builds on 25+ years of experience
Allows forest managers to assess carbon implications of
forest management: increase sinks, reduce sources
Available at carbon.cfs.nrcan.gc.ca
Outline
Background
Motivation
Methods
Results
Discussion
Conclusions
23
Reported and excluded area by category
Reported area is approximately constant over time (98.7 to 101.3% of 25 year avg. 186.6 Mha)
Excluded area is approximately constant over time (94.7 to 106.3% of 25 year avg. 45.5 Mha)
24
4: nd_pre_1990 <60 yrs (excluded)
5 and 7: nd_post_1990 (excluded)
6: u_nd_post_1990 (re-entered after ND)
3: ND_pre_1990 (re-entered after 60 yrs)
1: h_pre_1990
2: harvest_post_1990
Reported
Excluded
Implementation (preliminary results)
Sum of reported and excluded fluxes equals the NIR (excluding HWP)
25
Implementation (preliminary results)
Contribution of fluxes in stands affected after (Jan 1st 1990) by harvest
(left) or natural disturbances with more than 20% mortality (right).
26
Implementation (preliminary results)
Contribution of fluxes in stands clear-cut harvested prior to 1990 (left)
27
Implementation (preliminary results)
Contribution to fluxes of lands affected by stand-replacing ND (SRND)
over 60 years ago, or by partial ND with more than 20% mortality that
have reached pre-disturbance biomass (left) and lands affected by
SRND prior to 1990 but less than 60 years ago (right).
28
Implementation
Sum fluxes into reported and excluded categories.
Excluded categories will still be shown in NIR for transparency but
reported values will be the basis for entries into CRF tables for FL-FL
29
30
Implementation (preliminary results)
Sum of reported and excluded fluxes equals the NIR (excluding HWP)
31
FL-FL Reported Emissions from NIR 2016
are not correlated with human activities.
32
Revised Reported Emissions
are more closely correlated with human activities.
33
FL-FL
Revised Reported Emissions plus HWP emissions
are also closely correlated with activities.
34
FL-FL and HWP
Outline
Background
Motivation
Methods
Results
Discussion
Conclusions
35
DiscussionWhy do lands subject to ND more than 60 years ago re-enter the reporting?
Sustainable forest management is based on the concept that the amount of timber
harvested annually does not exceed the amount of regrowth. The total area of forest on
which the timber yield estimation and other forest management objectives are based is
larger than the area that has historically been harvested in Canada. In some areas of
Canada forests are being cut that have never before been harvested (although they will
have been disturbed previously by natural disturbances).
Thus the “managed forest” land area that is part of the sustainable management planning
process is larger than the area of forest previously harvested in Canada. The “managed
forest” area also includes lands that are never permitted to be harvested because they
serve other ecological functions such as riparian buffers, conservation areas and parks.
Timber rotations in the managed forests of Canada typically are 60 years or longer. Thus
stands that are 60 years or older are included in the management planning process. To
reflect that these lands are part of the managed forest we report emissions and removals
in stands that have been subject to stand-replacing wildfire 60 or more years ago.
36
DiscussionWhy the 20% mortality threshold for partial natural disturbances?
Stands affected by partial natural disturbances post 1990 are excluded from the reporting
only if the annual disturbance caused at least 20% mortality of aboveground biomass.
Disturbances with less than 20% mortality are considered part of forest dynamics.
Why re-entry when pre-disturbance biomass carbon stocks are reached?
Given the large area affected by insect disturbances, a criterion must be defined that
allows stands to re-enter the reporting. We selected above-ground biomass carbon
stocks because these are more readily monitored than total ecosystem carbon stocks.
Why re-entry when forest management follows natural disturbance?
Emissions associated with salvage logging, site preparation, slash burning or planting of
lands previously affected by natural disturbances must be reported. Therefore, forest
management activities on stands excluded from the reporting bring these stands back
into the reporting and emissions and subsequent removals are reported following salvage
logging.
37
Next Steps
Implementation and analyses are ongoing.
While a spatially-explicit modelling approach is not required, it may
increase “transparency” by documenting areas that are temporarily
excluded from the reporting.
Developing capability to implement spatially-explicit approach based
on next generation tools (moja.global).
Goal is to incorporate annual time series of disturbance maps