Comment Log Display

Mary Nichols, Chair
California Air Resources Board
Sacramento, CA 95812
Fax: (916) 322 - 3928


Re:	CARB Consideration of the California Climate Action Registry
Forest Protocols

Dear Chair Nichols and other members of the Air Resources Board:

I am writing your board to clarify some of the scientific and
technical issues related to the proposed California Climate Action
Registry Forest Protocols that appear to have been raised in
discussions leading up to the California Air Resources Board’s
deliberations on endorsement of the Forest Protocols.  I do so as
a scientist that has been involved in studying the issue of carbon
stores in forests for over 20 years.  During this time I have
published scores of peer-reviewed papers on this subject,
developed models of the processes involved, taught undergraduate
and graduate level classes, presented findings in national and
international scientific conferences and symposia as well as
public and government briefings, and been involved in the
development of national level research plans to study carbon
dynamics.  I am considered to be an expert in this arena and my
advice has been sought out by fellow scientists, government
agencies (state and federal), private land owners, consultants,
NGO’s and many others. In fact I was asked to provide guidance on
the Forest Protocols when they were initially being developed.

Below I list some important points regarding specific issues that
appear to have been raised. 

Carbon Sequestration by Younger versus Older Forests

It is very disappointing to find that arguments are still being
made that younger forests are better for climate mitigation than
older ones.  The mistaken basis for this argument is that younger
forests store carbon at faster rates than older forests.  There is
a grain of truth to the assertion that forests at a relatively
young age do have the potential to take up more carbon than older
forests.  But it is also true that forests younger than this
optimum age also take up less carbon.  Indeed immediately after
disturbance very young forests are releasing carbon as the dead
material caused by the disturbance (including timber harvests)
decomposes. Averaged over the entire period between disturbances,
the average flow into a forest equals the amount going out as long
as the same type of disturbance is repeated. This finding has been
repeatedly demonstrated in scientific examinations of this issue. 
The key is therefore not the rate of carbon uptake or 
release at any particular time, but the average amount stored over
time. I am not aware of a single scientific study in which the
average carbon store of a forest disturbed by clear cut harvesting
at a long interval is smaller than one disturbed at a shorter
interval.  Not a single study, and I just performed a literature
search on this very issue.  In addition to the interval between
disturbances, another important factor is the amount of carbon
removed by each disturbance.  Timber harvest, clear cutting in
particular, removes more carbon from the forest than any other
disturbance (including fire).  The result is that harvesting
forests generally reduces carbon stores and results in a net
release of carbon to the atmosphere.        
  
Another mistaken notion is that the Forest Protocols should focus
on rates of uptake and not changes in stores or stock changes. 
Scientists refer to these rates of carbon uptake and release as
fluxes. One must measure all the positive and negative fluxes to
understand the overall balance (much like in a bank balance in
which one must account for all the sources of income and expenses
for it to make sense). Simple mathematics tells us that as long as
all the relevant fluxes in and out of the forest are measured the
answer will be the same as if the changes in stocks are measured. 
The only difference is that measuring changes in stocks is far
easier and cheaper than accounting for all the fluxes. Scientists
measure fluxes to understand the mechanisms, but there is no need
to do this to determine the net change in carbon stores. A net
increase in stores is related to a positive flux into the forest,
a net decrease a negative flow out to the atmosphere, and no
change means the flows in and out are equal.  Both methods are
scientifically valid. 

Accounting for Wood Products 

In the Forest Protocols wood products are treated as an optional
carbon store.  I believe this is completely appropriate for
several reasons.  While it is true that some of the carbon
harvested from a forest is stored for a period of time it is not
the case that this material is stored forever.  Similar to other
forest-related pools, it is the balance of inputs versus outputs
that determines whether the wood products pool is increasing or
decreasing.  Not all harvested carbon results in storage into
longer term pools.  A considerable amount, estimated by the
guidelines to be 40%, is released to the atmosphere during
manufacturing and initial use. The remaining amount suffers losses
during use from fires, decomposition, and other factors.  We know
this because about half the wood products that are produced today
are used to replace the ones that have been in use.  I believe the
Forest Protocols addresses these issues adequately by providing
reasonable conversion factors, manufacturing losses, and product
life-spans that are based on previous peer-reviewed scientific
studies.  

Setting aside the specifics of how forest products could be
tracked, there are several reasons to make forest products
optional at this time.  First, is that even when this store is
included it only comprises a small fraction of the total forest
system stock of carbon. Again, based on a recent literature
review, less than 20% of the total forest system carbon story is
held in forest products.  The average fraction is likely less than
10%.  Second, unlike carbon in the forest itself, it is 
impossible to specifically account for where forest products end
up. Therefore there is no way to confirm the carbon stores are
actually present.  At least with a forest one can visit the actual
site of storage. Third, it is difficult to demonstrate the new
forest products meet additionality requirements: some of the new
material replaces old material and hence there is no real
additionality.  Granted the new harvest may help to maintain
current stores in forest products and that is accounted for under
the proposed Forest Protocols.   Fourth, the project supplying the
raw material has a limited ability to control the various products
that are produced and how and where they are used, which means
that the exact contribution to forest products pools is highly
uncertain.  At best the average storage rates can be computed
until a better way (probably incurring a great deal of expense) to
track the actual uses and life-span of products is developed.  

Use of default biomass coefficients

While it would be ideal if one could directly measure all the
carbon in a forest this is not practical at this time.  Instead
one must relate the size of the trees and other items to the
amount of carbon they store.  By making very detailed measures of
dimensions of each object (e.g., each tree) one can compute
volumes and coupling that with measurements of carbon content per
unit volume of each object one can very precisely determine carbon
stores in many kinds of forest pools. Unfortunately that would be a
very expensive process.  A more economical approach is to develop
biomass equations from a subsampling of trees or other objects. 
However, this too is has considerable expense and requires
technical training.  For those unable to develop or afford project
specific biomass equations, the Forest Protocols provide default
biomass regression equations that are reasonable and sound.  These
default equations were developed by respected and leading
scientists in the field of forest inventory (Richard Birdsey,
Linda Heath, Jennifer Jenkins and David Chojnacky) and were based
on a nationwide literature search using many thousands of diameter
measurements from a wide selection of many North American tree
species.   The equations were peer-reviewed, published by the USDA
Forest Service, and have become a national standard for scientific
study.

I see benefits other than economic ones in using the standardized
default equations.  It places everyone on equal footing and allows
for standardized checking of results. While the absolute carbon
store may be systematically over- or underestimated by these
equations, these biases are greatly reduced when the net change in
stocks is considered.  I see nothing whatsoever preventing
landowners from developing site specific biomass equations that
are more accurate than the default ones. The only restriction is
that the equations are approved by a third-party certifier, a step
that is essential to assure a credible program.   

Use of growth and yield models 

At the start of any project, it is logical to project the
potential increases in carbon stocks.  Projects unable to at least
predict a positive increase in carbon stores should not be
considered viable. Projections are ideally based on results from 
similar kinds of projects, but given the early stages of forest
carbon management, these data rarely exist.  A viable alternative
is use models to estimate potential project benefits. The Forest
Protocols specify a number of timber growth and yield models
including CACTOS (California Conifer Timber Output Simulator),
CRYPTOS (California Conifer Timber Output Simulator), FVS (Forest
Vegetation Simulator), SPS (Stand Projection System), VFP (Visual
Forester Professional), and FREIGHTS (Forest Resource Inventory
Growth, and Harvest Tracking System). I will not comment on the
merits of these specific models, however, I do note they were
pre-approved by the California Climate Action Registry and the
California Department of Forestry and Fire Protection which would
seem to be the appropriate institutions to conduct a model
evaluation and approval. If the models have a shortcoming it is
that they are largely focused on the live part of the forests and
do not include the other forest carbon pools. Still it is unlikely
that forests will increase overall carbon stores if the tree stores
are decreasing; therefore these models are a logical starting
point. 

As with other aspects of the Forest Protocols, projects are given
flexibility to develop their own projection models so long as they
have been reviewed by technically competent peers, are
parameterized for the specific conditions of the project, are used
within the scope for which they were developed and evaluated, and
are clearly documented.  Frankly I do not understand why anyone
would trust a model that was not reviewed, was parameterized for a
different set of conditions, and used for purposes it was not
developed for or was not documented. That would be completely
illogical. The Protocols also correctly point out that a
sensitivity analysis should be performed and that the models
should be periodically reviewed. Clearly it would be impossible to
understand any model unless one understands the various
uncertainties associated with it.  Periodic review is required
because models change as does the science they are based upon. 
The Forest Protocol requirements of annual reporting and direct
sampling of forest carbon (over ten year intervals) ensure that
the model projections are compared with ground-level data. By
coupling models and data one can more accurately forecast future
changes in carbon stores. Besides, the measured changes in carbon
stores are what actually happened, projections just what might
have happened. 


Requiring Confidence Level be Determined

While it is true carbon is carbon, not all carbon stores projects
are equally credible. There are two facets to this issue.  The
first is whether the project plan itself is viable. The Forest
Protocols deal with this issue by requiring information on the
location, climate, likely disturbances, longevity, proposed
activity and other factors that might influence the storage of
carbon.  Projects failing to meet these requirements should not be
considered viable.  The second is that those potentially viable
projects demonstrating actual increases in carbon stocks should
have more value than ones that do not.  As projects are likely to
use a range of sampling methods, the Forest Protocols correctly
uses the degree of statistical confidence to modify the estimate
of carbon stocks. These are used as deductions to provide a
conservative estimate of the most likely carbon store in a
project.  
This is entirely appropriate given underestimating stores causes
less potential environmental damage than overestimating the
stores. While this approach emphasizes the effect of sampling
errors (there are other kinds that are not considered), it is a
completely rigorous and technically sound way to factor in the
quality of the carbon store estimate. Given the sliding scale of
deductions the managers of a project can decide if the gains in
carbon related to reducing uncertainty outweigh the costs of
increased sampling.  Therefore this sliding scale discount
approach provides flexibility to landowners while ensuring a high
level of confidence in forest carbon estimates.   


Thank you for taking the time to consider these comments
concerning several scientific and technical aspects of the
California Climate Action Registry Forest Protocols.  I hope my
input clarifies several potential misunderstanding and leads you
toward the logical decision of endorsing the Forest Protocols as a
voluntary early action measure.

Sincerely,

Mark E. Harmon
Richardson Chair and Professor 
Forest Science