Environmental Science & Policy 1 (1998) 313±327
The Kyoto Protocol: provisions and unresolved issues relevant to
land-use change and forestry
Bernhard Schlamadinger *, Gregg Marland 1
Environmental Sciences Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831-6335, USA
Abstract
This paper describes the relevant text of the Kyoto Protocol and its implications for land-use change and forestry (LUCF)
activities and addresses some of the technical issues that merit further consideration and clari®cation before the treaty comes
into force. Although the phrasing of the Protocol is sometimes ambiguous and the opportunities limited, the Protocol does
provide for some selected forest-related activities to be used to meet national commitments for the reduction of greenhouse gas
emissions. To implement the forest-related portions of the Protocol, most importantly: (1) a clear de®nition for the word
`reforestation' is required, (2) contradictory wording in Article 3.3 needs to be clari®ed to establish how credits are to be
measured, (3) further thought should be given to the sentence in Article 3.7 which provides that countries with a net carbon sink
in LUCF in 1990 cannot include emissions from land-use change in their 1990 baseline, whereas countries with a net carbon
source in LUCF can include those emissions in their 1990 baseline, (4) the rules and baseline issues for joint implementation and
the clean development mechanism need to be clari®ed and (5) inclusion of additional forest management activities needs to be
considered. # 1998 Elsevier Science Ltd. All rights reserved.
Keywords: Forestry; Carbon sinks; A€orestation; Reforestation; Deforestation; Carbon accounting; Kyoto Protocol
1. Introduction
1.1. The Kyoto Protocol
Increasing anthropogenic emissions of greenhouse
gases to the atmosphere, observations that the concentrations of greenhouse gases in the atmosphere are
increasing and the expectation that these increases may
lead to deleterious changes in the climate have long
been recognized as a potential threat to the global environment (IPCC, 1996a,b). An important step in
addressing this threat was taken with adoption of the
United Nations Framework Convention on Climate
Change (UNFCCC, 1992). This Convention was
drafted in Rio de Janeiro in 1992 and has now been
* Corresponding author. Tel.: +1-423-574-7322; Fax: +1-423-5742232; E-mail: [email protected]
1
This paper is intended as a technical analysis of those portions of
the Kyoto Protocol that are concerned with land-use change and forestry. It is based on scienti®c studies supported by the US National
Science Foundation, The US Department of Energy, and the
Australian Science Foundation.
000-0000/98/$ - see front matter Published by Elsevier Science Ltd.
PII: S 1 4 6 2 - 9 0 1 1 ( 9 8 ) 0 0 0 1 6 - 1
rati®ed by 174 countries. The UNFCCC succeeded in
articulating global concern about the continuing
increase in greenhouse gas emissions and in encouraging developed countries (as listed in Annex I of the
Convention) to limit their emissions of greenhouse
gases, but it did not include binding commitments to
confront these emissions. Recognizing the inadequacy
of these commitments, the Conference of the Parties
under the Convention convened for the ®rst time in
Berlin in 1995 and decided that a strengthening of the
commitments was necessary and that this should be
pursued through a protocol or other legal instrument.
The Third meeting of the Conference of Parties convened in December, 1997, in Kyoto, Japan to try to
adopt a Protocol to the UNFCCC that would provide
binding commitments for countries to limit their greenhouse gas emissions and with speci®ed penalties for
those countries that did not meet their commitments.
After an arduous 11 days of negotiations, the Kyoto
Protocol was accepted by the delegates to the
Conference.
The Kyoto Protocol is now open for signature and,
as of 11 June 1998 had been signed by 40 countries.
314
B. Schlamadinger, G. Marland / Environmental Science & Policy 1 (1998) 313±327
The Protocol must subsequently be rati®ed and will
enter into force 90 days after being rati®ed by 55
countries which represent at least 55% of the carbon
dioxide emissions for 1990 of those countries listed in
Annex I of the UNFCCC (The Kyoto Protocol actually provides in Annex B a list of countries that di€ers
slightly from the list in Annex I of the UNFCCC).
The Kyoto Protocol achieved some of its objectives
and fell short on others. It prescribes limiting the emissions of 6 greenhouse gases or groups of gases and it
de®nes quantitative commitments for emissions reductions for the countries listed in Annex B of the
Protocol. It does not establish commitments for
countries not listed in Annex B (developing countries)
and it does not prescribe penalties for failure to meet
the commitments. The Protocol leaves many issues to
be resolved at a later date. If successful, the Protocol
would lead to a net reduction of 5.2% in emissions of
greenhouse gases from developed countries. For the
most important of the greenhouse gases (carbon dioxide, methane and nitrous oxide) emissions reductions
are de®ned with respect to emissions rates in 1990 and
the commitments are de®ned in terms of the average
of emissions in the 5-year period 2008±2012 (represented as 2008/2012 in this text). Bolin (1998) points
out that the objectives of the Protocol are modest in
terms of what may be needed to restrain climate
change, but that they are an important ®rst step.
Carbon emissions from land-use change and forestry
were a critical part of negotiations leading to drafting
of the Kyoto Protocol and only in the morning hours
following the last formal day of the Conference was a
compromise reached that was accepted by all Parties.
Most Parties, in fact, left the Conference aware of
some of the residual problems in the text that would
need to be addressed in the processes of rati®cation
and implementation.
1.2. Land-use change and forestry
Land-use change and forestry (LUCF) activities
have often been cited as (i) historically, and currently,
providing a net source of anthropogenic emissions of
carbon dioxide to the atmosphere (i.e. from deforestation) (Dixon et al., 1994; Houghton, 1996) and (ii) potentially providing opportunities to reduce net carbon
dioxide emissions to the atmosphere or to increase the
net uptake of carbon from the atmosphere (IPCC,
1996b). For one or both of these reasons, many observers felt that it was desirable to include LUCF activities in a binding treaty limiting greenhouse gas
emissions (Baltimore Conference, 1997). Several
options for inclusion of LUCF were discussed before
and at the Kyoto conference, including total omission
of LUCF from the Protocol. Many considerations,
including data uncertainty, veri®ability, equity among
countries, etc. (also including factors like confusion,
misunderstandings, misconceptions, politics and lack
of negotiating time) played a role in forming the
Protocol text. Appendix A of this paper contains
Articles 3.3, 3.4 and 3.7 of the Kyoto Protocol, all of
which discuss LUCF and all of which are examined in
this paper. The Kyoto Protocol was drafted under
severe time pressures and with con¯icting national
objectives and the phrasing in a number of places
leaves room for interpretation. The discussion below is
based on our interpretation of the Protocol language
and points out places where we feel the language is
particularly unclear.
The core of this paper is in two parts:
. Section 2 provides some observations on the
Kyoto Protocol and what is speci®cally included or
not included regarding land-use change and forestry.
. Section 3 elaborates on some problems, questions
and issues to be resolved where the intent of the
Protocol is not clear or the application of Protocol
language could create perverse outcomes. In Section 3
we also try to suggest ways in which the Protocol
might be interpreted and/or implemented so as to
minimize confusion, minimize adverse outcomes, and
maintain the spirit under which the Protocol was
drafted.
2. Some observations on the Kyoto Protocol regarding
land-use change and forestry
2.1. Land-use change and forestry activities are included
but are limited to speci®ed activities
The Kyoto Protocol establishes that human-induced
land-use change and forestry activities, limited to a€orestation, reforestation and deforestation, since 1990,
``shall be used to meet the commitments'' (Article 3.3).
Land-use change and forestry is, however, not
included among the sectors/source categories listed in
Annex A for inclusion in calculating ``aggregate
anthropogenic carbon dioxide equivalent emissions''
(Article 3.1). In other words, it appears that emissions
from LUCF are not included in the 1990 emissions
baseline for a country but can be used to meet the
country's commitments in the ®rst commitment period
(2008/2012). We note here that Article 3.1 makes reference to the gases listed in Annex A of the Protocol but
fails to mention the sectors/source categories listed in
Annex A. Given that these sectors/source categories
are not mentioned anywhere else in the Protocol, we
interpret that reference should also be made to the sectors/source categories.
Article 5.1 would seem to require that a country
should report all emissions by sources and removals by
sinks and Article 3.4 clearly calls for a report of the
B. Schlamadinger, G. Marland / Environmental Science & Policy 1 (1998) 313±327
`level of carbon stocks in 1990', but the net of emissions and uptake in land-use change and forestry is
not included in the 1990 baseline on which national
obligations are based (with an exception outlined in
Article 3.7 and discussed in Section 3.3). The bottom
line is that `direct human-induced' a€orestation reforestation and deforestation since 1990, measured as
`veri®able changes in carbon stocks in each commitment period', are not used to de®ne commitments but
shall be used to meet emissions commitments. A key
word in Article 3.3 is `veri®able'. The framers of the
Kyoto Protocol were clearly concerned with the ability
of countries to document the amount of carbon stored
in sinks and explicitly require that only `veri®able'
changes can be credited.
We interpret a `change in carbon stocks' in the ®rst
commitment period to be the di€erence between the
stocks on 31 December 2012 and on 1 January 2008.
This interpretation is reinforced by a conclusion
drafted by the eighth session of SBSTA (Subsidiary
Body for Scienti®c and Technological Advice to the
UNFCCC) during its meetings in June 1998
(UNFCCC, 1998b) which says: ``[SBSTA] understands
the meaning of Article 3.3 in the Kyoto Protocol to be
as follows: the adjustment to a Party's assigned
amount shall be equal to veri®able changes in carbon
stocks during the period 2008 to 2012 resulting from
direct human-induced activities of a€orestation, reforestation and deforestation since 1 January 1990.
Where the result of this calculation is a net sink, this
value shall be added to the Party's assigned amount.
Where the result of this calculation is a net emission,
this value shall be subtracted from the Party's assigned
amount''.
2.2. The text clearly calls for assessing the changes in
carbon stocks
Article 3.3 states that emissions and sinks from
land-use change and forestry activities will be
``measured as veri®able changes in stocks in each commitment period'' and Article 3.4 asserts that each
Annex I country shall provide data to ``establish its
level of carbon stocks in 1990 and to enable an estimate to be made of its changes in carbon stocks in
subsequent years''. This intent to focus on carbon
stocks is reinforced by the conclusions drafted by
SBSTA in June 1998, as quoted above (UNFCCC,
1998b). There had been considerable discussion (see,
for example, Apps et al., 1997; Winjum et al., 1998)
whether carbon emissions from land-use change and
forestry should be measured as the net carbon ¯ux
between forest reservoirs and the atmosphere (assessment of atmospheric ¯uxes), as is done for fossil-fuel
combustion, or whether the net change in carbon
stocks (assessment of changes in carbon stocks) is a
315
more appropriate choice for estimating net carbon
emissions from renewable reservoirs. Assessment of
carbon exchange with the atmosphere had been proposed in 1996 by an IPCC (Intergovernmental Panel
on Climate Change) expert group drafting guidelines
for reporting national estimates of emissions related to
harvested wood products but was referred for reconsideration by the 12th IPCC plenary session in Mexico
City (September, 1996).
The choice of whether to assess carbon stock
changes or carbon ¯uxes makes considerable di€erence
for Parties who buy or sell harvested wood products.
For example, the assessment of changes in carbon
stocks places emissions from burning biofuels in the
account of a Party practicing non-sustainable forest
harvest rather than in the account of the Party burning
the harvest as a fuel. If biofuels are produced sustainably, both producer and consumer would report zero
net carbon emissions under a stock-change approach.
Quantitative examples for di€erent forestry and wood
utilization scenarios can be found in Apps et al. (1997)
and Winjum et al. (1998) illustrate the di€erent
approaches by looking at data for selected countries.
With the wording of the Protocol the choice seems to
have been made in favour of a stock-change approach.
2.3. Harvested wood products are not mentioned
The Protocol does not mention harvested wood products. Short of careful crafting of additions to the
IPCC (1997a,b) Guidelines, it appears that the
Protocol does not permit credit for a carbon sink
when carbon is stored in long-lifetime wood products.
If the IPCC or the UNFCCC can ®nd appropriate
phrases to include this carbon sink within the constraints of Protocol language, it appears that (in order
to be compatible with the forestry sector calculations)
a carbon source or sink in wood products will have to
be represented by the change in carbon stocks in products rather than as some measure of the ¯ux to the
atmosphere from oxidation of wood products (see
Section 2.2). Note also that once a land area enters the
Kyoto accounting framework by virtue of a€orestation
or reforestation since 1990, it appears likely that full
accounting of the carbon in the forest and in harvested
products from these `Kyoto forests' will be necessary
(see also Section 3.9 of this paper). A minimum
requirement would be to show that claiming credits for
carbon sinks in `Kyoto forests' does not result in a
decrease of carbon stocks in wood products.
2.4. More `human-induced activities' can be added
Article 3.4 of the Kyoto Protocol states that the meeting of the Parties ``... shall, at its ®rst session or as soon
as practicable thereafter, decide upon modalities, rules
316
B. Schlamadinger, G. Marland / Environmental Science & Policy 1 (1998) 313±327
and guidelines as to how, and which, additional humaninduced activities [besides a€orestation, reforestation,
and deforestation] related to changes in greenhouse gas
emissions and removals by sinks in the agricultural soil
and land-use change and forestry categories shall be
added to, or subtracted from, the assigned amount for
Parties included in Annex I''. Any such decision shall
apply in commitment periods after the ®rst, except that
a Party can choose to apply them in the ®rst commitment period. This provision leaves open the possibility
that credit could be obtained for other carbon sinks, but
current limitation to the three listed activities re¯ects
concern that credits be limited to those activities that
can be best documented (`veri®able') and that credits be
awarded for actions and not circumstances (for
example, fertilization of forests by increasing atmospheric CO2 is not a direct human-induced activity that
could be credited against future commitments).
Some human-induced activities that might be added
are mentioned in Section 3.10 of this paper and a more
comprehensive list can be found in UNFCCC (1998a).
3. Some problems, questions and issues to be resolved
3.1. What exactly does `since 1990' mean?
Article 3.3 of the Protocol speci®es that emissions
from sources and removals by sinks from LUCF activities will be ``... limited to a€orestation, reforestation
and deforestation since 1990...''. In the cases of a€orestation and reforestation, does this mean that only
projects that were initiated since 1990 can be counted,
or does it also include carbon uptake since 1990 in
projects that were initiated prior to 1990? Our interpretation is that the intent is the former (projects initiated since 1990), but if this is what the delegates had
in mind, supporting documents to the Protocol should
make this interpretation clear.
It is also not straightforward whether `since 1990'
includes the year 1990. It initially appeared to us that
the intent here was to indicate times following the end
of 1990. De®nitions of `since' in Webster's unabridged
dictionary include both contexts in which 1990 would
be or would not be included. SBSTA, at its eighth session in Bonn in June 1998, has now concluded that
`since 1990' should be interpreted to begin on 1
January 1990 (UNFCCC, 1998b).
The phrase `since 1990' creates a more serious problem. While deforestation in 2008 (for example) could
release the bulk of the contained carbon to the atmosphere in 2008, reforestation in 2008 would provide a
continuing carbon sink for many years thereafter, but
with an annual carbon stock change per hectare that is
very much smaller than the annual stock change per
hectare from deforestation. This important di€erence
in the time rate of carbon ¯ux is discussed further in
Sections 3.2 and 3.3. In Appendix C an example is
given to illustrate the relative sizes of land under a€orestation and deforestation needed in order for the two
e€ects to cancel out in the ®rst commitment period.
3.2. Should reforestation be considered and how?
The Kyoto Protocol permits consideration of
changes in carbon stocks due to `a€orestation, reforestation, and deforestation' without providing de®nitions
for these three words. The words `a€orestation' and
`reforestation' (but not `deforestation') are de®ned in
the glossary of the Intergovernmental Panel on
Climate Change (IPCC) Guidelines for National
Greenhouse Gas Inventories (IPCC, 1997a) and since
the Protocol refers to these Guidelines for estimating
national emissions, we assume that these de®nitions
are those that should apply. The words a€orestation
and deforestation do not create any severe problems
and the intent of the Protocol seems consistent with
conventional de®nitions of these two words. It
appears, however, that the intent of this phrase was to
limit credits to cases where there is a clear change in
land use and where we can clearly verify that this
change in land use has occurred. Accepting this link to
the IPCC Guidelines, the framers of the Protocol
appear to have used `a€orestation' to identify establishment of forest where forest had not historically
existed whereas `reforestation' appears to have been
used to suggest establishment of forest where there
was forest previously but the land had been converted
to other uses. It is not clear how long a time lapse
between forest harvest and forest regrowth is required
in order to qualify as reforestation.
In Appendix B we reproduce some published de®nitions of `reforestation' and these make it clear that
many foresters would include the natural or assisted
regrowth of trees immediately following harvest to be
included in `reforestation'. To the UN Forestry and
Agriculture Organization (FAO), for example, reforestation includes immediate replanting or regrowing of
forests following harvest or a natural disturbance such
as ®re. Using the FAO de®nition of reforestation, the
text of the Kyoto Protocol would thus imply that no
emissions are to be associated with forest harvest and
yet emissions reductions could be credited for the
reforestation that follows. Using the IPCC de®nition
of reforestation, guidance is required to distinguish
between forest management and the potentially linked
activities of deforestation and reforestation. The distinction is particularly important in the initial phase of
the Protocol where it is possible that deforestation
could occur prior to the ®rst commitment period while
reforestation and regrowth would occur, and provide
sequestration credits, during a commitment period.
B. Schlamadinger, G. Marland / Environmental Science & Policy 1 (1998) 313±327
If credits could be obtained for reforestation (as
de®ned by the FAO), then harvest should lead to debits (emissions). Otherwise clearing old-growth forest
with the purpose of replanting fast-growing monocultures could lead to signi®cant credits from the beginning of such an undertaking, despite net emissions to
the atmosphere. If credits are obtained for reforestation (as de®ned by the IPCC Guidelines) the converse
process needs to be identi®ed as `deforestation' rather
than harvest and there needs to be a clear understanding of the length of time that land resides in other
land uses before it quali®es for `reforestation'. The
question is the point at which a forest or land-use
change activity causes the land to enter the accounting
framework (i.e. to become a Kyoto forest). Clearly the
Protocol does not intend, at this point, to provide
credits for activities that increase the carbon density of
forests without changing the land use, nor does it
intend to count emissions where forest degradation
leads to lower carbon density in the forest so long as
the land maintains its identity as forest. It appears that
the framers of the Protocol intended that land should
enter the accounting framework at the point that there
is a change in land use to or from forest land. This
intent is not completely consistent with the traditional
de®nitions of `reforestation' nor is the philosophy consistent with inclusion of credits for reforestation without debits for harvest.
We suggest that there are four options for confronting the ambiguity of what is meant by `reforestation'.
Although the ®rst option is the one that is consistent
with literal reading of the Protocol language and the
IPCC de®nition of reforestation, the other options are
suggested by other de®nitions of `reforestation' or
could be implemented with more liberal interpretations
of the Protocol text.
Option 1: Include, in supporting documents to the
Protocol, a precise de®nition for `reforestation' that mentions a minimum time period
for non-forest land use, a de®nition that
di€ers from traditional de®nitions by
excluding replanting or revegetation activities that immediately follow harvest or
natural disturbance.
Option 2: Use something like the FAO de®nition of
reforestation and interpret reforestation to
include any harvest which necessarily precedes it.
Option 3: As in option 2 but, in addition, include
regrowth from harvests prior to 1990.
Option 4: Omit reforestation from the accounting.
Option 1 is based on the assumption that the delegates drafting the Kyoto Protocol had in mind a de®nition of reforestation similar to that given in the
317
IPCC Guidelines. A more explicit and more precise
de®nition is needed to make clear that this is land that
was deforested earlier with no expectation that the
land would be returned to forest. A minimum time
period between deforestation and reforestation could
be mentioned. A de®nition for the purposes of the
Protocol could be as follows:
Reforestation: establishing forests on lands which
have, historically, previously contained forests but
which have been converted to some other use. This
other use must have prevailed for at least 20 (or other
number to be chosen) years, or, alternatively, the other
use can be shorter if the land has been counted as
deforested within a commitment period under the
Kyoto Protocol.
One could choose a number greater or less than 20
years depending on the intent to include less or more
land under `reforestation', but if the number becomes
less than perhaps 15 years, this would allow more
opportunity for forests that are clear-cut before the
®rst commitment period to be reforested for credits
during a commitment period.
Note that `establishing' is used instead of `planting'
(the original term in the glossary to the IPCC
Guidelines), because this avoids a discussion of
whether human assisted natural revegetation is to be
included. By adding the phrase ``or, alternatively,
...under the Kyoto Protocol'', a piece of land once
classi®ed as deforested in a commitment period can be
counted as reforested, without creating a loophole for
emissions to the atmosphere. Essentially this re¯ects a
full carbon accounting for land once reported under
the Kyoto Protocol (see also Section 3.9).
As a minimum, it should be clear that accelerated
deforestation prior to the ®rst commitment period, followed by reforestation within the commitment period,
is in violation of the spirit of the Protocol.
With this de®nition for reforestation we could then
de®ne.
Deforestation: conversion of forest land for other
land use.
Option 2 would mean that all managed forests harvested and replanted after 1990 would become part of
the accounting. Harvest would be considered by subtracting the removed biomass carbon from the forest
carbon stocks. For the sake of this discussion, consider
a country with one hectare of land (with 100 tC ha ÿ 1)
cleared and reforested each year and with the rate of
regrowth at 1 tC ha ÿ 1. Thus, in 1991 the system of
carbon accounting would include one hectare harvested and the same hectare regrowing; in 1992 one
hectare would be harvested and 2 hectares would be
regrowing, etc. Starting in 1991, a country would get a
debit (carbon emission) of 100 tons each year. The
credit would be 1 tC in 1991, 2 tC in 1992, 3 tC in
318
B. Schlamadinger, G. Marland / Environmental Science & Policy 1 (1998) 313±327
1993 and so on, with the total balance in 1991 being a
net emission of 99 tC, in 1992 of 98 tC etc. A key
point here is that harvest can occur in a single year
while reforestation occurs over decades. If the same
total amount of land is reforested that is deforested
each year after 1990, there will be a net loss of carbon
stocks in the accounts reportable to the Protocol for
many years, until the system reaches equilibrium (see
also Appendix C).
Option 3 would solve the problems presented in
option 2 by the di€erent time rates of carbon ¯ux
during harvest and reforestation. This approach would
recognize the long-term balance of a sustainable forest
management system but might contravene the statement on `direct human-induced activities since 1990' in
the Protocol. The accounting in option 3 would thus
involve all managed forests (this would, in essence, be
the `gross-net approach' initially proposed by New
Zealand in Kyoto) and this has important implications
regarding equity between di€erent national commitments.
Option 4 would avoid the problems encountered in
options 2 and 3 by excluding reforestation from the
accounting for carbon sinks.
The problem with the phrasing in the Protocol is
that it does not explicitly de®ne reforestation but
seems to rely on a non-traditional (and not suciently
detailed) de®nition of the word. Most of the problems
can be solved by clari®cation of the apparent intent of
the word `reforestation' as suggested under `option 1'
above.
We note that clari®cation of the de®nition of `a€orestation' or `forest' may be needed to make clear
whether establishment of fast growing tree plantations
with short rotation period is intended to be included.
3.3. Countries with net carbon emissions from land-use
change and forestry in 1990 are treated di€erently than
countries with a net carbon sink in 1990
The last sentence of Article 3.7 contains the
phrases `land-use change and forestry' (®rst part of
the sentence), and `land-use change' (last part of the
sentence). The text in Section 3.3.1 opposite was written under the assumption, that there was a purposeful
distinction and that the words `and forestry' were purposefully included in the ®rst part of the sentence but
not included in the last part of the sentence. In section
3.3.2 opposite we discuss the implications if there was
no distinction intended and the words `and forestry'
were either inadvertently omitted in the last part of the
sentence or inadvertently included in the ®rst half of
the sentence.
3.3.1. Purposeful distinction between ``land-use change''
and ``land-use change and forestry''
Article 3.7 of the Protocol provides that countries
with a net source of greenhouse gas emissions from
LUCF in 1990, while still bound by Articles 3.3 and
3.4 of the Protocol, can include their ``aggregate
anthropogenic carbon dioxide equivalent emissions by
sources minus removals by sinks in 1990 from land-use
change'' in the 1990 baseline from which their assigned
amounts are calculated. It was perceived that this sentence was necessary to avoid inequities between
countries and to make it possible for countries with
net emissions from land-use change in 1990 to meet
part of their commitments through a reduction of
these net emissions.
The phrase ``emissions by sources minus removals
by sinks in 1990 from land-use change'', if taken literally, would include the e€ects of land-use change initiated prior to 1990 on the carbon emissions inventory
in 1990. For example, carbon uptake in 1990 in a forest stand that was planted (a€orested) in 1950 would
be included. It would, perhaps, have been more appropriate to write `deforestation' instead of `land-use
change' (the latter including all of a€orestation, reforestation and deforestation) in this sentence, because
the main motivation of this provision in Article 3.7
was apparently to confront the special situation of
countries with deforestation in 1990 (e.g. Australia).
This intent could be retained (and the confusion of
having to account for 1990 impacts of all previous
a€orestation and reforestation avoided) by interpreting
the latter part of this sentence to read ``...emissions by
sources minus removals by sinks in 1990 from land-use
change (i.e. deforestation)''. The net result would be
that emissions from deforestation would be included in
the national baseline for those countries for which
LUCF activities were a net source of carbon emissions
in 1990.
In Appendix D we show with a simple calculation
how application of the last sentence of Article 3.7 to
two hypothetical countries (A and B) does indeed produce an equitable balance between two countries with
comparable emissions from fossil fuels and comparable
e€ort to improve land-use management over the interval 1990 to 2008/2012, when one had emmisions from
deforestation in 1990 and the other did not.
However, although the ®nal sentence of Article 3.7
creates equity for a country with deforestation and an
overall net carbon emission from LUCF in 1990, the
same principles create problems for a country with
both deforestation and a larger carbon sink elsewhere
in its forests in 1990. The need for this sentence arises
ultimately from the inherent di€erence between a€orestation and reforestation (or other activities that generate carbon sinks) on the one hand and deforestation
on the other hand, i.e. the di€erence in time rate of
B. Schlamadinger, G. Marland / Environmental Science & Policy 1 (1998) 313±327
carbon ¯ux introduced in Section 3.1 above. Carbon
emissions originating from a constant rate of deforestation from 1990 to 2010 will result in a constant ¯ow
of debits2. Carbon uptake occurring at a constant rate
from 1990 to 2010 will create no credits (sink) in 2010
to the extent that it results from a€orestation initiated
prior to 1990, or to the extent that it is due to activities that generate carbon sinks but cannot be included
under a€orestation or reforestation. In other words, if
we move from 1990 to the ®rst commitment period
with no change in behavior, most of the debits for
emissions from LUCF remain constant while most of
the credits suddenly disappear because of the `since
1990' requirement, the di€erence in the time dependent
behavior of the two types of ¯uxes, and the limitation
of sinks to a€orestation and reforestation. In
Appendix D we show with another simple calculation
how the last sentence in Article 3.7 impacts a hypothetical country (country C) with both deforestation and
a larger carbon sink elsewhere in its forests in 1990. It
becomes obvious that the sentence in Article 3.7 does
not apply to country C and thus country C is at a disadvantage compared to country B as far as commitments for fossil-fuel-related carbon emissions are
concerned.
The last sentence of Article 3.7 in the Kyoto
Protocol, if interpreted literally, thus seems to provide
bene®ts to countries for which LUCF was a net source
of carbon emissions in 1990 whereas these bene®ts are
not available to countries for which LUCF was a net
carbon sink, despite emissions from deforestation, in
1990.
3.3.2. No intended distinction between ``land-use
change'' and ``land-use change and forestry''
This inequity is changed signi®cantly if we interpret
the intent of the sentence such that both parts of the
sentence are the same and `and forestry' is either
included both times or omitted both times.
If we interpret this sentence in Article 3.7 with the
assumption that `land-use change and forestry' is a
uni®ed phrase and that `and forestry' should apply to
both parts of the sentence, then the sentence provides
bene®t to the same parties but the bene®t is smaller
for countries with sinks in managed forests. This is
because not only land use change, but the net of all
other carbon stock changes in forestry in 1990 will be
2
We recognize that this is not strictly true, because not all organic
matter is oxidized in the year of deforestation, and there will likely
be continuing oxidation of carbon from the soils and slash of areas
deforested in earlier years. Nonetheless, a large fraction of oxidation
occurs in the ®rst year following deforestation and if the majority of
carbon is oxidized within 20 years then there will be constant debits
in the 1990 base year and, from 20 years of deforestation ``since
1990'', in the 2008/2012 commitment period.
319
used to calculate the national commitment. Appendix
D provides an example for hypothetical country D.
If we interpret the last sentence in Article 3.7 to indicate that a distinction is being made between landuse change and forestry versus land-use change alone
and that this sentence was intended to apply uniformly
and consistently to only land-use change, (or, to deforestation alone, Section 3.3.1 opposite) then the bene®t
of the sentence is shared by all countries with deforestation in 1990. This is because other carbon sinks in
forests will not disqualify a country with deforestation
in 1990 from using the emissions from deforestation to
de®ne its commitments. Although this interpretation of
Article 3.7 seems to o€er the most equitable treatment
of all countries experiencing deforestation in 1990, it
might be judged politically unacceptable because it
does serve to raise the 1990 baseline (with respect to
the literal interpretation), against which commitments
are calculated, for all countries with deforestation in
1990. Appendix D shows with parenthetical numbers
how application of this interpretation a€ects the commitments of countries C and D noted above.
In summation, while recognizing the problem of
raising the baseline for many countries, we suggest
that the most equitable interpretation of the last sentence in Article 3.7 may be as follows: ``Those Parties
included in Annex I for whom land-use change (i.e.,
deforestation) constituted a net source of greenhouse
gas emissions in 1990 shall include in their 1990 emissions base year or period the aggregate anthropogenic
carbon dioxide equivalent emissions by sources minus
removals by sinks in 1990 from land-use change (i.e.,
deforestation) for the purposes of calculating their
assigned amount''.
If developing countries eventually take on emission
limitation commitments, there are likely to be more
countries with both deforestation and a€orestation or
other forest sinks in the base year, so that the issues
described in this section could become increasingly important.
3.4. The emissions baseline: should it be based on 1 year
or a longer period?
For a country with net emissions from LUCF in
1990, a literal interpretation of Article 3.7 provides
that land-use change net emissions in 1990 are to be
incorporated into the baseline. Emissions in 1990 can
be considerably di€erent than emissions in 1989 or
1991. For example, wood removals in Germany in
1990 were almost twice those in 1989 or 1991 (FAO,
1998). It would be desirable if, for LUCF, the initial
source or sink could be determined for a base period
(e.g. 1988 to 1992) rather than for a single year. This
is the approach taken by the Protocol for de®ning the
320
B. Schlamadinger, G. Marland / Environmental Science & Policy 1 (1998) 313±327
commitment in terms of the mean of emissions over a
®ve-year period.
3.5. Leakage from forestry projects under the `clean
development mechanism' can circumvent project
objectives
Consider a project in which an Annex I country
acquires emission credits for a forest-based mitigation
project in a non-Annex I country. If the sequestered or
protected carbon is released to the atmosphere after
the project lifetime (perhaps 20 or 30 years), which
party would assume responsibility for this net emission? (Marland et al., 1998). Under the Protocol, nonAnnex I countries would have no commitment for
emissions limitations so could not be held responsible
for emissions even though they had been host to certi®ed emission reduction credits.
One option is to assign this net emission to the piece
of land from which it occurred. If another CDM project wanted to use that land it would have to take on
that net emission as an initial debit. Another possibility is for non-Annex I countries to accept emissions
commitments that cover only LUCF activities. For
example, Articles 3.3, 3.4 and 3.7 (last sentence only)
could eventually be extended to cover (some) nonAnnex I countries (perhaps as a prerequisite for participation in LUCF projects), thereby also providing
additional participation in the Protocol commitments
by developing countries. Some non-Annex I countries
have considerable net emissions from deforestation
and some have a great potential for a€orestation or
reforestation.
3.6. Countries may be allowed to include or exclude
certain parts of LUCF
The last sentence of Article 3.4 leaves the decision to
individual countries whether to report the carbon
sources and sinks from `additional human-induced' activities during the ®rst commitment period. Countries
with increased carbon stocks in the respective categories would likely tend to report this, whereas
countries with a net decrease of carbon stocks might
choose not to report. Failure to report carbon sources
under this article is likely to represent a small loophole
in the Protocol in that no such additional activities
have yet been endorsed and the option of not reporting applies only to the ®rst commitment period.
3.7. Do a€orestation, reforestation and deforestation,
include carbon sequestered in forest soils?
The Protocol does not explicitly say whether the
change in carbon stocks in forest soils and forest litter
can or shall be accounted for in a€orestation, refores-
tation, and deforestation activities. Presumably the
controlling word here will be `veri®able' and changes
in carbon stocks in forest soils could be used to meet
the commitments to the extent that they can be veri®ed. However, there would be no incentive to report
and verify negative stock-changes, e.g. from deforestation. It should be possible to require that if carbon
gains are eligible for credits as a result of a€orestation
or reforestation, a reporting and veri®cation of carbon
gains or losses in the forest soil would have to be
included. As a minimum, when credits are claimed for
increasing the stock of carbon in the above-ground
portions of a forest, it should be demonstrated that no
loss of soil carbon has occurred. The Protocol speci®cally reserves ``changes in greenhouse gas emissions by
sources and removals by sinks in the agricultural
soils... categories'' for possible inclusion under the additional activities of Article 3.4. (see Section 3.10).
Note that greenhouse gas emissions other than CO2
from agricultural soils are already included in Annex
A of the Protocol.
3.8. The Protocol asks countries to establish the `levels
of carbon stocks in 1990'
Article 3.4 calls on Annex I countries to `provide
data to establish their level of carbon stocks in 1990'.
Such data are not required to establish or evaluate
compliance with the commitments under the Kyoto
Protocol, nor are they asked for in the current IPCC
Guidelines. Performance under the Protocol requires
only an accounting of the carbon stocks on 1 January
2008 and 31 December 2012 (or their di€erence) and
so far only for forests impacted by a€orestation, reforestation, or deforestation since 1990 (Kyoto forests).
3.9. A€orestation and reforestation: a full carbon
accounting is required in subsequent commitment periods
There is a distinction made between forests a€ected
by Article 3.3 of the Protocol (Kyoto forests) and
other managed forests not undergoing land-use change
(for a pictorial representation see IGBP Terrestrial
Carbon Working Group, 1998). Any stock change in
the other forests appears to be left out from the
accounting of national emissions commitments.
However, according to Article 3.3, a stock change in
areas a€orested or reforested since 1990 can be credited toward a country's commitments. It appears that
for these Kyoto forests, a full carbon accounting has
to be continued in the second and in future commitment periods, even if there are no further increases in
stocks. If this is not the case, a Party could, for
example, claim credit for increasing the stocks in
standing trees from 0 to 100 tC per ha, but then avoid
B. Schlamadinger, G. Marland / Environmental Science & Policy 1 (1998) 313±327
reporting a subsequent decrease in carbon stocks,
perhaps resulting from increased harvest levels.
3.10. Which other land-use activities or land-use changes
should be considered to be added under the protocol?
321
Any such project provides a reduction in emissions
by sources, or an enhancement of removals by
sinks, that is additional to any that would otherwise
occur.
Other ``additional human-induced activities in the
agricultural soil and land use change and forestry categories'' can be added under the Kyoto Protocol
(Article 3.4). This can occur at the ®rst meeting of the
Parties after entry into force of the Protocol. The
requirement that changes in carbon stocks be veri®able, is likely to play a major role in the acceptance of
other activities for o€setting emissions. All of these
additional activities would be accounted for on an
activity by activity basis. Activities that might be
accepted include (see UNFCCC, 1998a for a more
complete list):
(I) Changes in forest carbon stocks (compared to a
baseline scenario) due to changes in forest management practices. Examples are: selective logging instead
of clear-cutting, reduced-impact logging, and lengthening the rotation period. Such options could bring
about a temporary improvement of the carbon balance. Comparison with a baseline scenario is important because forest carbon stocks might change even
without additional human-induced activity (see also
Schlamadinger and Marland, 1997).
(II) Changes in soil carbon stocks on agricultural
land due to changes in tillage or other farming
practices. Again, a baseline scenario will be needed
to establish the change in carbon stocks due to
additional, human-induced activities.
Note that whereas Article 3.3 mentions `direct
human induced land-use change and forestry activities', the word direct is not included in the phrasing of
Article 3.4.
Article 6 explicitly includes sources and sinks in any
sector of the economy but does not limit LUCF activities to a€orestation, reforestation, and deforestation
since 1990 (as does Article 3.3). This leaves open the
question, whether other human-induced LUCF activities (that would ful®ll also the `additionality' requirement) can be credited in JI projects. Although the text
does not speci®cally limit JI to the same activities as
listed for domestic activities, such a limitation makes
sense in practical terms. While activities under Articles
3.3 or 3.4 will increase the assigned amount of the
country where the activities are carried out
(UNFCCC, 1998b) (no matter who carries them out,
either the country itself or a JI donor from another
country), this is not the case for activities not accepted
under Articles 3.3 or 3.4. Articles 3.10 and 3.11 of the
Kyoto Protocol provide that any transfers of JI emission reduction units, while increasing the assigned
amount of one party, will decrease the assigned
amount of another party by the same amount. Thus,
any LUCF projects not admitted by Articles 3.3 and
3.4 will merely have the same e€ect as emissions
trading outlined in Article 17 without increasing the
sum of the assigned amounts of the participating
countries (see also Greenpeace, 1998).
3.11. What forestry activities can be included under joint
implementation
Article 3.3 clearly limits the activities that can be
accounted in meeting commitments from domestic
sources and sinks of carbon. Article 6 of the Protocol
addresses Joint Implementation (JI) between Annex I
countries by saying (Article 6.1):
In de®ning the Clean Development Mechanism
(CDM), the Kyoto Protocol provides incentives for
Annex I countries to ``assist Parties not included in
Annex I in achieving sustainable development and in
contributing to the ultimate objective of the
Convention'' by developing emissions reduction activities within the non-Annex I countries. Article 12.3(b)
of the Protocol states:
For the purpose of meeting its commitments under
Article 3, any Party included in Annex I may transfer to, or acquire from, any other such Party emission reduction units resulting from projects aimed
at reducing anthropogenic emissions by sources or
enhancing anthropogenic removals by sinks of
greenhouse gases in any sector of the economy...
Parties included in Annex I may use the certi®ed
emission reductions accruing from such project activities to contribute to compliance with part of
their quanti®ed emission limitation and reduction
commitments under Article 3, as determined by the
Conference of the Parties serving as the meeting of
the Parties to this Protocol
and (Article 6.1 (b))
3.12. What forestry activities can be included under the
clean development mechanism?
and (Article 12.5 (c)).
322
B. Schlamadinger, G. Marland / Environmental Science & Policy 1 (1998) 313±327
... Reductions in emissions that are additional to
any that would occur in the absence of the certi®ed
project activity.
The only other requirement for these certi®ed emission reductions is that they provide ``real, measurable,
and long-term bene®ts related to the mitigation of climate change'' (Article 12.5(b)). However, unlike in
Article 6 for Joint Implementation, there is no speci®c
mention in Article 12 of `removals by sinks'. At this
point it is not clear whether this omission was intentional. Even if the omission of sinks from the text of
Article 12 was intentional, this text should still allow
for the inclusion of emission reductions in forestry, for
example by reducing or preventing deforestation. As in
the case of Joint Implementation projects, additionality
is required for CDM projects and there is no speci®c
limitation whether sources and sinks in LUCF (if
included at all) are limited to the activities a€orestation, reforestation and deforestation since 1990.
Carbon sequestration in existing forests is not explicitly excluded for CDM projects. For CDM projects a
clear baseline (see also Section 3.13) for mitigation
projects would have to be established in order to calculate the `additional C sequestration', and the condition of veri®ability would have to be observed.
3.13. What are the baselines against which emissions
from direct, human-induced activities are to be
measured?
It appears that the baseline for measuring the contribution of domestic a€orestation, reforestation and
deforestation to meeting emissions commitments is a
zero baseline (i.e., no a€orestation, no reforestation,
no deforestation). Land-use change and forestry activities are not included in the emissions calculation for
1990 on which commitments are based and there is no
`additionality' required. An exception is made for
countries for which there were net emissions from
LUCF activities in 1990, and for these countries the
``aggregate anthropogenic carbon dioxide equivalent
emissions by sources minus removals by sinks in 1990
from land-use change'' forms part of the 1990 baseline.
Credits against commitments are limited to a€orestation, reforestation and deforestation during the commitment period, but which of these activities, for
countries that qualify under Article 3.7, are included in
the 1990 baseline depends on the interpretation chosen
for Article 3.7 of the Protocol (see Section 3.3 of this
paper).
For activities outside of the national borders,
whether pursued under the Clean Development
Mechanism or as Joint Implementation activities with
other Annex I countries, there is a requirement of
additionality. That is, credits can be passed on to
another country only when they are in excess of what
would have accrued without the speci®c project (baseline scenario). As an example, it appears that if abandoned agricultural land returns to forest by natural
regeneration, CO2 emissions credits could be claimed
for reforestation of domestic land but there would be
no emissions credits to be traded through either a JI
or CDM agreement.
3.14. There is a logical inconsistency in the language of
Article 3.3
In Article 3.3, the main body of the ®rst sentence
reads: ``The net changes in greenhouse gas emissions
by sources and removals by sinks resulting from direct
human-induced land-use change and forestry activities... shall be used to meet the commitments under
this Article...''. This speci®cation that improvements in
LUCF activities can be used to meet emissions commitments is limited by two additional clauses. One
clause in this sentence speci®es that these net changes
in emissions will be ``measured as veri®able changes in
carbon stocks in each commitment period''. To appeal
to the fundamentals of calculus, the `changes in carbon
stocks' will be the ®rst time derivative of the stocks
and will have dimensions of tons per year (t yr ÿ 1). On
the other hand, the ``net changes in greenhouse gas
emissions... and removals...'' will then necessarily be
the second time derivative of the stocks and will have
dimensions of tons per year per year (t yr ÿ 1 yr ÿ 1). In
other words, ``changes in stocks during the commitment period'' cannot be used to measure ``the net
changes in... emissions... and removals...''.
The words `net change' at the beginning of Article
3.3 should be ignored to assure clarity of the Article.
The conclusions issued by SBSTA in June, 1998
(UNFCCC, 1998b) concur that Article 3.3 should be
interpreted in this way. The same problem occurs in
the middle of Article 3.4.
4. Summation
In the area of land-use change and forestry the
Kyoto Protocol tried to do several things. Recognizing
that LUCF has very large potential as a CO2 source
and signi®cant potential as a CO2 sink, the Protocol
tried to include consideration of LUCF in order to
minimize net emissions from these activities and thus
from human-induced activities as a whole. In addition
to their desire to reduce net CO2 emissions to the atmosphere, it appears that those who drafted the
Protocol were guided by 3 principles: (1) to provide
credits against emissions commitments only for activities that could be reliably measured and veri®ed, (2) to
B. Schlamadinger, G. Marland / Environmental Science & Policy 1 (1998) 313±327
provide credits against emissions commitments only
for direct human-induced activities and not for advantageous circumstances and (3) to treat the various
Parties to the Protocol in an equitable way. This
proved to be a daunting challenge; not the least
because forestry is based on a cyclic, potentially renewable and sustainable, system of growth, harvest, oxidation and renewal and because these processes
proceed at widely di€erent rates.
The Protocol tried to optimize on some parts of the
total system; veri®able, human-induced changes in carbon stocks in the biosphere. It was perhaps inadvertent, but we believe that the choice to focus on
assessment of changes in carbon stocks provides incentive for ecient use of renewable biofuels and, perhaps, durable wood products (see Apps et al., 1997).
But hoping to optimize the system by optimizing on a
part of the system is likely to produce perversities or
inequities somewhere. It appears that these have inevitably occurred, some because of time pressures in
drafting of the Protocol, some because of focus on the
3 objectives listed above and some, perhaps, because
of political motivations. Our hope in compiling this
document is to identify the worst of the oversights and
perversities and to shed light on how some of them
might be ®xed and how others might be minimized
even if not eliminated. The Protocol does appear to be
a very good beginning.
Acknowledgements
This paper is intended as a technical analysis of
those portions of the Kyoto Protocol that are concerned with land-use change and forestry. It is based
on scienti®c studies supported by the U.S. National
Science Foundation (agreement number ATM9711602) and the Global Change Research Program
of the Environmental Sciences Division (Oce of
Biological and Environmental Reseach) of the
U.S. Department of Energy. B. Schlamadinger was
supported by the ``Erwin SchroÈdinger Auslandsstipendium'' program of the Austrian Science
Foundation (FWF) under contract number J 1467
BIO. We would like to thank 4 anonymous reviewers
and numerous other experts for valuable comments on
earlier versions of this manuscript.
Appendix A
Paragraphs addressing sinks in the Kyoto Protocol
(UNFCCC, 1997)
Article 3, Paragraph 3: The net changes in greenhouse gas emissions from sources and removals by
323
sinks resulting from direct human-induced land-use
change and forestry activities, limited to a€orestation,
reforestation and deforestation since 1990, measured
as veri®able changes in stocks in each commitment
period, shall be used to meet the commitments under
this Article of each Party included in Annex I. The
greenhouse gas emissions by sources and removals by
sinks associated with those activities shall be reported
in a transparent and veri®able manner and reviewed in
accordance with Articles 7 and 8.
Article 3, Paragraph 4: Prior to the ®rst session of
the Conference of the Parties serving as the meeting of
the Parties to this Protocol, each Party included in
Annex I shall provide, for consideration by the
Subsidiary Body for Scienti®c and Technological
Advice, data to establish its level of carbon stocks in
1990 and to enable an estimate to be made of its
changes in carbon stocks in subsequent years. The
Conference of the Parties serving as the meeting of the
Parties to this Protocol shall, at its ®rst session or as
soon as practicable thereafter, decide upon modalities,
rules and guidelines as to how, and which, additional
human-induced activities related to changes in greenhouse gas emissions by sources and removals by sinks
in the agricultural soils and the land-use change and
forestry categories shall be added to, or subtracted
from, the assigned amount for Parties included in
Annex I, taking into account uncertainties, transparency in reporting, veri®ability, the methodological
work of the Intergovernmental Panel on Climate
Change, the advice provided by the Subsidiary Body
for Scienti®c and Technological Advice in accordance
with Article 5 and the decisions of the Conference of
the Parties. Such a decision shall apply in the second
and subsequent commitment periods. A Party may
choose to apply such a decision on these additional
human-induced activities for its ®rst commitment
period, provided that these activities have taken place
since 1990.
Article 3, Paragraph 7: In the ®rst quanti®ed emission limitation and reduction commitment period,
from 2008 to 2012, the assigned amount for each
Party included in Annex I shall be equal to the percentage inscribed for it in Annex B of its aggregate
anthropogenic carbon dioxide equivalent emissions of
the greenhouse gases listed in Annex A in 1990, or
the base year or period determined in accordance
with paragraph 5 above, multiplied by ®ve. Those
Parties included in Annex I for whom land-use
change and forestry constituted a net source of greenhouse gas emissions in 1990 shall include in their
1990 emissions base year or period the aggregate
anthropogenic carbon dioxide equivalent emissions by
sources minus removals by sinks in 1990 from landuse change for the purposes of calculating their
assigned amount.
324
B. Schlamadinger, G. Marland / Environmental Science & Policy 1 (1998) 313±327
Appendix B
De®nitions for a€orestation, reforestation and
deforestation, and for forests
(1) FAO Forest Resources Assessment, 1990 (concepts, de®nitions and methodology):
Forests are de®ned as ``ecological systems with a
minimum of 10% crown coverage of trees and/or bamboos, generally associated with wild ¯ora, fauna and
natural soil conditions, and not subject to agricultural
practices''.
Deforestation refers to ``change of land use from
forest to other land use or depletion of forest crown
cover to less than 10%''. Changes within the forest
class which negatively a€ect the stand or site and, in
particular, lower the production capacity, are termed
forest degradation. Thus degradation is not re¯ected in
the estimates of deforestation.
(2) Temperate and Boreal Forest Resources
Assessment 2000, terms and de®nitions, UN-ECE/
FAO (1997a, p. 3):
Forest: land with tree crown cover (or equivalent
stocking level) of more than 10 percent and area of
more than 0.5 ha. The trees should be able to reach a
minimum height of 5 m at maturity in situ. May consist
either of closed forest formations where trees of various
storeys and undergrowth cover a high proportion of the
ground; or of open forest formations with a continuous
vegetation cover in which tree crown cover exceeds
10%. Young natural stands and all plantations established for forestry purposes which have yet to reach a
crown density of 10% or tree height of 5 m are included
under forest, as are areas normally forming part of the
forest area which are temporarily unstocked as a result
of human intervention or natural causes but which are
expected to revert to forest.
Includes: forest nurseries and seed orchards that constitute an integral part of the forest; forest roads, cleared
tracts, ®rebreaks and other small open areas within the
forest; forest in national parks, nature reserves and
other protected areas such as those of special environmental, scienti®c, historical, cultural or spiritual interest; windbreaks and shelterbelts of trees with an area of
more than 0.5 ha and a width of more than 20 m.
Rubberwood plantations and cork oak stands are
included.
Excludes: land predominantly used for agricultural
practices.
(3) State of the World's Forests (FAO, 1997b, pp.
173±174):
A€orestation/rea€orestation: the establishment of a
tree crop on an area from which it has always or very
long been absent. Where such establishment fails and
is repeated, the latter may properly be termed rea€orestation.
Reforestation: establishment of a tree crop on forest
land.
Deforestation (developed countries): change of forest
with depletion of tree crown cover to less than 20%.
Deforestation (developing countries): change of forest
with depletion of tree crown cover to less than 10%.
(Changes within the forest class, e.g. from closed to
open forest, which negatively a€ect the stand or site
and, in particular, lower the production capacity, are
termed forest degradation and are considered apart
from deforestation.)
Forest (developed countries): land with tree crown
cover (stand density) of more than about 20% of the
area. Continuous forest with trees usually growing to
more than about 7 m in height and able to produce
wood. This includes both closed forest formations
where trees of various storeys and undergrowth cover
a high proportion of the ground, and open forest formations with a continuous grass layer in which tree
synusia cover at least 10% of the ground.
Forests (developing countries): ecosystem with a minimum of 10 percent crown cover of trees and/or bamboos, generally associated with wild ¯ora, fauna and
natural soil conditions and not subject to agricultural
practices.
(4) IPCC Revised 1996 Guidelines for National
Greenhouse Gas Inventories (IPCC, 1997a), Reporting
Instructions (Vol. 1), Glossary, p. 1 and 15:
A€orestation: planting of new forests on lands
which, historically, have not contained forests. These
newly created forests are included in the category
`Changes in Forest and Other Woody Biomass Stocks'
and in the Land Use Change and Forestry module of
the emissions inventory calculations.
Reforestation: planting of forests on lands which
have, historically, previously contained forests but
which have been converted to some other use.
Replanted forests are included in the category
`Changes in Forest and Other Woody Biomass Stocks'
in the Land Use Change and Forestry module of the
emissions inventory calculations.
A somewhat diverging de®nition, used in a di€erent
context (to explain `plantations') is in the Revised
IPCC Guidelines for National Greenhouse Gas
Inventories (IPCC, 1997b) Reference Manual (Vol. 3),
p. 5.14:
Plantations are forest stands that have been established arti®cially, to produce a forest product `crop'.
They are either on lands that previously have not supported forests for more than 50 years (a€orestation) or
on lands that have supported forests within the last 50
years and where the original crop has been replaced
B. Schlamadinger, G. Marland / Environmental Science & Policy 1 (1998) 313±327
with a di€erent one (reforestation) (see also Brown et
al., 1986).
(5) See Lund (1998) for a comprehensive overview
of de®nitions for a€orestation, reforestation deforestation and forests.
Appendix C
Example describing the inherent di€erence between
a€orestation and deforestation with respect to the time
rate of change of carbon stocks
In this example we estimate the area of land under
a€orestation that is required for a country to o€set 1
ha yr ÿ 1 of deforestation (with loss of 100 tC ha ÿ 1).
The country in our example is assumed to have a net
carbon sink in LUCF in 1990, so that the last sentence
of Paragraph 3.7 does not apply. We assume that forest stands in an a€orestation program grow at 1 tC
ha ÿ 1 yr ÿ 1 and that the land being deforested carries
100 tC ha ÿ 1.
The emissions in the commitment period from
deforestation amount to 100 tC annually. In order to
o€set this carbon source, an average of 100 ha of
land with `new forest', growing at 1 tC ha ÿ 1 yr ÿ 1,
are needed in the commitment period. This requirement is ful®lled if 100 ha of a€orested land are accumulating carbon through the duration of the
commitment period. This requirement is also ful®lled
if the area a€orested is increasing with time so that
325
there are 80 hectares in the year 2008, 90 hectares in
the year 2009, etc. with 120 hectares in the year
2012. In other words, an a€orestation program involving the planting of trees on 10 ha each year, starting
on 1 January 2001, could exactly o€set deforestation
of 1 ha land per year in the commitment period. If,
however, the a€orestation program started ®ve years
later (1 January 2006), then an annual planting of 20
hectares would be required.
This suggests that the land put into an a€orestation
program each year has to be 1 order of magnitude (a
factor of perhaps 10 to 20, depending on how long
before the commitment period the a€orestation program begins) greater than the land deforested annually
in order to gain positive carbon credits. This factor
will become lower, the longer the accounting system is
in e€ect, and eventually approach 1. For the growth
rate and carbon density chosen here, the area being
occupied with a€orestation projects in a given year has
to be 2 orders of magnitude greater than the land
deforested in the same year. This factor will not
change over time.
Appendix D
What Article 3.7 of the Kyoto Protocol does and what it
could do (see Section 3.3): some examples
We illustrate the impact of the last sentence of
Article 3.7 (see Appendix A) on the emissions commitments of 4 hypothetical countries in order to show the
Table 1. Article 3.7 of the Kyoto Protocol applied to hypothetical countries A, B, C and D described in
Appendix D. The numbers in brackets for countries C and D are derived using the suggested interpretation of
Article 3.7 (see end of Section 3.3)
Country A
Country B
Country C
Country D
10
0
0
0
10
2
0
2 source
10
2
3
1 sink
10
2
1
1 source
1990 Baseline for target
10
12
10 (12)
11 (12)
2008/12 Target w/o a€., def.
9.5
11.4
9.5 (11.4)
10.45 (11.4)
2008/12
2008/12
2008/12
2008/12
0
1
1 sink
0a
1
0
1 source
0a
1
0
1 source
3 sourcea
1
0
1 source
1 sourcea
10.5
10.4
8.5 (10.4)
9.45 (10.4)
1990
1990
1990
1990
Fossil fuel emissions
Deforestation
Sink in managed forests
LUCF total
Deforestation
New a€orestation
Deforestation + new a€.
Sink in managed forests
Target for fossil fuel emissions
a
Not counted under the Kyoto Protocol and thus irrelevant for the calculations.
326
B. Schlamadinger, G. Marland / Environmental Science & Policy 1 (1998) 313±327
di€erential impact that this sentence has on countries
depending on the details of their emissions by sources
and removals by sinks in LUCF and on the interpretation of the details of this sentence.
Country A (see also column 2 of Table 1).
Emissions of carbon from fossil fuels are taken to
be 10 units in 1990 and the net carbon sink in LUCF
is close to zero in 1990.
Country B (see also column 3 of Table 1).
Emissions of carbon from fossil fuels are taken to
be 10 units in 1990 and LUCF provides a source of 2
units of carbon from deforestation in 1990.
We assume that by 2008/2012 both countries
improve by 1 unit their carbon balance in LUCF
through human-induced activities since 1990. Country
A does this via a€orestation since 1990 and country B
by reducing deforestation. Both countries have an
emission reduction target of 5%. Country A gets a
credit in 2008/2012 for its sequestration of 1 unit and
is thus allowed to increase its fossil-fuel carbon emissions from 10 to 10.5 units. Country B has a 1990
baseline of 12 units and the debit for LUCF activities
in 2008/2012 (deforestation) is 1 unit. This debit would
have been 2 units if country B had not reduced deforestation. The target for carbon emissions in 2008/2012
is thus 11.4 units (95% of 12) and, with the LUCF
debit of 1 unit, the fossil-fuel emissions target is 10.4
unit, almost the same as for country A.
Without the addition in Article 3.7 country B would
have had a 1990 baseline of 10 units and, with the 5%
reduction and 1 unit debit from LUCF, the reduction
target for fossil-fuel emissions would have been 8.5
units. The last sentence of Article 3.7 succeeds in creating equity between these two countries with similar circumstances and similar e€ort to reduce net emissions.
Country C (see also column 4 of Table 1).
Country C is as country B, except that it has in
1990 an additional carbon sink of 3 units in its managed forests. We assume that country C reduces its rate
of deforestation by 1 unit in 2008/2012 and that the
carbon uptake in its managed forests remains
unchanged. The 3 units of carbon removal by managed forests do not count toward meeting Kyoto commitments because they do not meet the requirements
of Article 3.3, but under the literal interpretation of
Article 3.7 they prevent country C from including the
2 units of emissions from deforestation in its 1990
baseline.
The 1990 baseline for country C is thus 10 units and
the 2008/2012 target is 9.5 units. There are direct
human-induced activities in LUCF in the commitment
period resulting in 1 unit per year of emissions from
deforestation so that fossil-fuel emissions have to be at
8.5 units in the commitment period. This is 2 units
lower than for country B despite similar circumstances
and similar e€ort to reduce net emissions from LUCF.
If we use the interpretation of Article 3.7 which is
suggested at the end of Section 3.3, the computation
for countries A and B will be unchanged but the computation for country C will be exactly as for country B
(see the parenthetic numbers under country C in
Table 1) because country C is now able to similarly include the emissions from deforestation in its baseline.
Country D (see also column 5 of Table 1).
Country D is as countries B and C except that there
is a carbon sink of 1 unit in managed forests in 1990.
Because there is a net carbon source in LUCF, country
D is able to include deforestation in its 1990 baseline.
We use country D to illustrate the consequences if the
last sentence of Article 3.7 is interpreted to include
`and forestry' in both parts of the sentence (see Section
3.3). In this case only the net of carbon emissions from
LUCF (emissions from deforestation minus sink in
managed forests) is used to de®ne the 1990 baseline so
that the target for 2008/2012 is 10.45 units (95% of
11). With 1 unit of deforestation continuing in 2008/
2012, country D has a target for emissions from fossil
fuels of 9.45 units in 2008/2012. This is intermediate
between countries B and C as country D is able to derive part of the bene®t provided by the literal interpretation of Article 3.7.
If we use the interpretation of Article 3.7 which is
suggested at the end of Section 3.3, the computation
for country D will be as shown in the parenthetic
numbers under column D in Table 1 and are the same
as calculated for countries A, B and C (parenthetic
number) which have similar circumstances and have
made similar e€ort to reduce net emissions from
LUCF.
References
Apps, M., Karjalainen, T., Marland, G., Schlamadinger, B., 1997.
Accounting system considerations: CO2 emissions from forests, forest products, and land-use change, a statement from Edmonton.
Http://www.joanneum.ac.at/IEA-Bioenergy-TaskXV.
Baltimore Conference, 1998. Biotic o€sets environment workshop,
®nal report of a workshop in Baltimore MD, 5±7 Sept. 1997,
Treiler and Associates Inc., Portland OR 97267-3513.
Bolin, B., 1998. The Kyoto negotiations on climate change: a science
perspective. Science 279, 330±331.
Brown, S., Lugo, A.E., Chapman, J., 1986. Biomass of tropical tree
plantations and its implications for the global carbon budget. Can.
J. Forest Res. 16, 390±394.
B. Schlamadinger, G. Marland / Environmental Science & Policy 1 (1998) 313±327
Dixon, R.K., Brown, S., Houghton, R.A., Solomon, A.M., Trexler,
M.C., Wisniewski, J., 1994. Carbon pools and ¯ux of global forest
ecosystems. Science 263, 185±190.
FAO, 1997a. Food and Agriculture Organization of the United
Nations. Temperate and Boreal Forest Resources Assessment
2000, Terms and De®nitions. UN-ECE/FAO, pp. 3.
FAO, 1997b. State of the World's Forests. ISSN 1020-5705. Rome.
FAO, 1998. Forest Products Data. FAOSTAT database at Http://
www.fao.org.
FAO Forest Resources Assessment, 1990. Http://faov02.fao.org:70/
0gopher_root%3a[fao.fra]def_uk.txt.
Greenpeace, 1998. Greenpeace analysis of the Kyoto Protocol.
Greenpeace Brie®ng Paper. UNFCCC sessions of the Subsidiary
Bodies, Bonn, June 2±12, 75 pp.
Houghton, R.A., 1996. Land-use change and terrestrial carbon: the
temporal record. In: Apps, M.J., Price, D.T. (Eds.), Forest
Ecosystems, Forest Management and the Global Carbon Cycle,
NATO ASI Series, vol. I 40. Springer-Verlag, Berlin, pp. 117±134.
IGBP Terrestrial Carbon Working Group, 1998. The terrestrial carbon
cycle: implications for the Kyoto Protocol. Science 280, 1393±1394.
IPCC, 1996a. Climate change 1995. The science of climate change.
Contribution of Working Group I to the Second Assessment
Report of the Intergovernmental Panel on Climate Change.
Cambridge University Press, NY 10011-4211, USA.
IPCC, 1996b. Climate change 1995. Impacts, adaptations and mitigation of climate change. Contribution of Working Group II to the
Second Assessment Report of the Intergovernmental Panel on
Climate Change. Cambridge University Press, NY 10011-4211,
USA.
IPCC, 1997a. Intergovernmental Panel on Climate Change. In:
Houghton, J.T. et al. (Eds.), Revised 1996 Guidelines for National
Greenhouse Gas Inventories, Greenhouse Gas Inventory
Reporting Instructions, vol. I. IPCC/OECD/IEA 1997, IPCC
WGI Technical Support Unit, Bracknell, UK.
327
IPCC, 1997b. Intergovernmental Panel on Climate Change. In:
Houghton, J.T. et al. (Eds.), Revised 1996 Guidelines for National
Greenhouse Gas Inventories, Greenhouse Gas Inventory
Reference Manual, vol. 3. IPCC/OECD/IEA 1997, IPCC WGI
Technical Support Unit, Bracknell, UK.
Lund, G., 1998. De®nitions of deforestation, a€orestation and reforestation. A Report prepared for the USDA Forest Service and
IUFRO 6.03.02. Http://home.att.net/~gklund/DEFpaper.html.
Marland, G., Schlamadinger, B., Feldman, D., 1998. Reforestation:
what happens when the JI project ends? In: Riemer, P.W.F.,
Smith, A.Y., Thambimuthu, K.V. (Eds.), Technologies for activities
implemented jointly. Pergamon, 1998.
Schlamadinger, B., Marland, G., 1997. A proposal for inclusion of
land-use change and forestry in a protocol to be adopted in
Kyoto. Revised version of Nov. 19, 1997. Available from the
authors (e-mail: [email protected]).
UNFCCC, 1992. United Nations Framework Convention on Climate
Change. Convention on Climate Change. UNEP's Information Unit
for Conventions (IUC), Http://www.unep.ch.
UNFCCC, 1997. Kyoto Protocol to the United Nations Framework
Convention on Climate Change. Document FCCC/CP/1997/7/
Add.1. Http://www.unfccc.de.
UNFCCC, 1998a. Methodological issues: issues related to land-use
change and forestry. Note by the UNFCCC Secretariat.
Document FCCC/SBSTA/1998/INF.1. 8th Session of the
Subsidiary Body for Scienti®c and Technological Advice, Bonn,
June 2±12. Http://www.unfccc.de.
UNFCCC, 1998b. Methodological issues: land-use change and forestry. Decision 1/CP.3, Paragraph 5(a), Document FCCC/SBSTA/
1998/CRP.3. 8th Session of the Subsidiary Body for Scienti®c and
Technological Advice, Bonn June 2±12. Http://www.unfccc.de.
Winjum, J., Brown, S., Schlamadinger, B., 1998. Forest harvests and
wood products: sources and sinks of atmospheric carbon dioxide.
Forest Sci. 44, 272±284.