Why We Disagree About Forest Carbon. I. Atkins Article on Oregon Wild Assertions

I am extremely appreciative of David Atkins’ effort to delve into the complicated area of forest carbon in this piece at TreeSource. I actually asked several scientists if they knew of a paper that outlined the different assumptions and conclusions, and discussed the area of disagreements.. maybe someone out there knows of one? Otherwise, you have to wonder why that would not be the most important kind of paper to write..Here are the areas of disagreement, here is why, and here is what we need to do to find out more. If we are going to use science in policy, wouldn’t that be the most important kind of investigation to conduct?
It’s not the universities’ job though, here’s what Acting Dean Davis of OSU says:

“Researchers often explore extremes of a subject on purpose, to help define the edges of our understanding; or other studies might only examine one aspect of an issue which in reality does not occur in a vacuum. It is important to look at the whole array of research results around a subject rather than using those of a single study or publication as a conclusion to a field of study.”

That might be difficult for policy makers to do unassisted, and I wouldn’t blame them for throwing up their hands at what appears to be an arena of scientist gladiators.

I think it falls to government to address these policy questions, through research designed as Peter Gluckman describes here.

But how could researchers come to such wildly different conclusions on the carbon effect of wood products? This led me to a series of interviews and multiple other sources to sort through a rabbit warren of questionable assumptions and conclusions in the OSU researchers’ paper.

Make assumptions with care

Did the scientific process break down in the review – or lack of review – given this paper and its assumptions and conclusions?

The problems that surfaced in the Law paper include:

* The quote used by Oregon Wild can’t be found in the references cited.
* The calculation used to justify doubling forest rotations assumes no leakage. Leakage is a carbon accounting term referring to the potential that if you delay cutting trees in one area, others might be cut somewhere else to replace the gap in wood production, reducing the supposed carbon benefit.
* The paper underestimates the amount of wildfire in the past and chose not to model increases in the amount of fire in the future driven by climate change.
* It assumes a 50-year half-life for buildings instead of the minimum 75 years the ASTM standard calls for, which reduces the researchers’ estimate of the carbon stored in buildings.
* It assumes a decline of substitution benefits, which other LCA scientists consider as permanent.
* It models just one species of insect to account for tree mortality when there are a variety of insect and diseases which impact forest carbon capture and storage. And the insect mortality modeled was unrealistic.
* The OSU scientists assumed wood energy production is for electricity production only. However, the most common energy systems in the wood products manufacturing sector are combined heat and power (CHP) or straight heat energy production (drying lumber or heat for processing energy) where the efficiency is often two to three times as great and thus provides much larger fossil fuel offsets than the modeling allows.
* The researchers claim to conduct a Life Cycle Assessment (LCA), but fail to use the international standards for conducting such analyses, without explaining this difference in methods.
* The peer reviewers did not include an LCA expert.
* The claimed significance of substantial carbon savings from delaying harvest and the large emission numbers from the forest products sector are undermined by all of the above.

I think the pros and cons of the different assumptions might be an interesting conversation to have on this blog. And we could perhaps use this blog, if scientists were interested, as a forum for extended peer review.

17 thoughts on “Why We Disagree About Forest Carbon. I. Atkins Article on Oregon Wild Assertions”

  1. Yep, some variables haven’t even been recognized yet, and some variables are outright ignored. “Beyond the scope of this study” often means that the ‘desired results’ would be in question if that important aspect were explored.

    Reply
  2. Thanks for an interesting post. On first reading I had the impression that the list of problems with the Law article came from the OSU Dean of Forestry rather than Atkins, whose Board of Directors includes a WSJ reporter, a former USFS chief, and a forest products marketer. It’s a good start for what can hopefully become a great conversation and a useful exposition, if not exactly the equivalent of an objective, peer-reviewed scientific article like the Law paper.

    Atkins, being a journalist, can publish whatever he wants unlike Dr Harmon, whose work must pass peer review. Atkins’ surprise at his failure to be able to draw Harmon, a scientist, into making broad policy pronouncements seems a little disingenuous, and perhaps reflects on the TreeSource pro-industry slant in the article.

    Among other unanswered questions the article raises one might point out that the problem of leakage is not at all a scientific phenomenon, but is political. Should their laws change to mirror the kind of longer rotation forestry the Law paper explores, timber imports from the countries which quickly filled the timber deficit left by the NW Forest Plan could evaporate overnight. Choosing to regard existing export arrangements as permanent is a political, rather than a scientific judgement, and deserving of in-depth and critical analysis.

    Another relevant point that can be drawn from the Atkins article is the much greater scale of resources backing research oriented to the industry point of view about forest carbon (including TreeSource itself) in contrast to Law who, since Harmon’s recent retirement, essentially works alone.

    Here’s hoping this discussion leads to the kind of peer-reviewable, permanent, long term on the ground study of the very young and highly uncertain science of the forest carbon cycle (including substitution) needed to apply Gluckman’s principles to this most timely discussion about the future of forest management.

    Reply
  3. Atkins paper does not add to the scientific conversation and muddies the public conversation.

    Atkin says:
    ” It boils down to whether they made reasonable assumptions. … The problems that surfaced in the Law paper include: … • It assumes a decline of substitution benefits, which other LCA scientists consider as permanent.”

    This is just an example of “he said. she said.”

    In fact, the scientists who have criticized high estimates of substitution benefits of wood products have given specific, well-reasoned support for their criticism, and many substitution proponents such as Atkin continue to stick by the flawed analyses.

    Law & Harmon conducted a literature review and concluded …

    Most LCA [life cycle analysis] studies rely heavily on wood product substitution for GHG benefits, and these have been grossly overestimated, with many ambiguous assertions that gloss over forest carbon dynamics; for example:
    • Biofuel emissions are assumed to be zero because they are balanced by net growth, yet this would depend on the state of the preceding forest system – they could be positive, neutral or negative;
    • Old forests are assumed to always be carbon sources, while young forests are always assumed to be carbon sinks, contrary to forest carbon dynamics findings;
    • Dead wood and soil carbon stores are either not included or assumed to be constant;
    • In one LCA, dead wood is not present in older forests, contrary to findings in the extensive ecological literature;
    • The wood product pool is assumed to be an increasing carbon stock over time.

    Substitution of more energy-intensive building materials with a less energy intensive one can, in theory, result in a fossil fuel offset; for example, when wood replaces a construction material with higher emissions (e.g., concrete or steel), the fossil CO2 emission avoided by choosing wood is credited as an offset. Thus, harvest of forest carbon and placement into buildings can impact the overall carbon balance of the forest sector [33,42]. However, several additional factors need to be considered. First, changes in the carbon stores of the forest ecosystem have to be considered relative to a base case that includes a lower level of harvests. As noted above, decreasing the interval between harvests, or increasing harvest intensity will lower the carbon store in the forest [9–11,31]; the question is whether stores in forest products combined with substitution offsets surpass losses from shorter rotations. Since the forest has a maximum carrying capacity, just the growth in carbon stores and offsets would seem to eventually exceed old forest carbon, although it could take centuries to happen, even using the most generous substitution effects. With more realistic substitution effects, it may never happen. In some cases, the amount of live and dead biomass in unharvested forests was grossly underestimated leading to an overestimation of the relative benefits of substitution. Second, in substitution effects calculations, it is often tacitly assumed that wood that is removed from forests and used in long-term wood products, specifically buildings, continues to accumulate infinitely over time. While building carbon stores have increased in many areas (e.g., the USA), this is largely because more forest area is being harvested and not because the harvest-related stores per harvest area are increasing. The trend that is being used as evidence of increasing building stores is based on the fact that because a greater area has been harvested, the total store has increased. This is not the same thing as the increase associated with a particular area of forest. A fixed per area basis is how substitution effects have largely been evaluated in the past, so arguing on an expanding area basis is inappropriate. The reason that wood products saturate is that housing and other wood products have a finite lifespan and are eventually replaced [43]. Although there can be some reuse of wood, essentially assuming an infinite lifespan or 100% reuse of wood products is completely unrealistic. Carbon is always lost as wood products are used or disposed of, which means release of CO2 to the atmosphere. Since long-term storage in forest products saturates over time (i.e., eventually does not increase), the effect of substituting wood for fossil fuel energy is also likely to saturate. Third, in most cases, the substitution offset was calculated based on the assumption that each time a house is to be built, the preference is for nonwood materials. This results in an estimate of the maximum substitution effect possible, but does not account for actual preferences for building materials. Granted, preferences vary by region and over time, but without accounting for these one cannot possibly estimate realistic substitution benefits. Fourth, current substitution accounting appears to violate a key principle of carbon offsets, namely permanence. In fact the ever-increasing substitution offset presented in these analyses appears to depend on impermanence of wooden buildings. Fifth, most, if not all, current analyses of substitution effects ignore the effects of additionality and whether wooden buildings are initially present. Given that many forests have already been harvested to produce wood products, replacing wooden buildings with more wooden buildings results in no additional substitution effect. Finally, these studies assume that it is a permanent benefit to GHG removal from the atmosphere. That is, they assume there is a continual increase in the carbon credit, and maintenance of a sustainable productive forest dedicated to providing substitutes for nonwood fuels and materials [44].
    These caveats all suggest that while there is likely to be some building material substitution effect that is valid, it is far lower than generally estimated and as subject to saturation as other forest-related carbon pools. In summary, the substitution effect appears to have been grossly overestimated. Substitution is an offset, not a store. Offsets depend on the use of appropriate accounting rules. Until rules such as permanence, additionality and leakage are followed, the values being presented in many analyses are not credible.

    Life cycle analysis (including substitution, proposed considerations)

    • Substitution of more energy intensive building materials with less energy intensive ones can in theory result in a fossil fuel offset, but important considerations suggest that the substitution effect is substantially lower than estimated, and is subject to saturation.

    Beverly Elizabeth Law & Mark E Harmon 2011. Forest sector carbon management, measurement and verification, and discussion of policy related to mitigation and adaptation of forests to climate change. Carbon Management 2011 2(1). http://terraweb.forestry.oregonstate.edu/pubs/lawharmon2011.pdf

    An alternative to increasing carbon stores within the forest is to harvest wood and store some of this carbon within wood products (Perez-Garcia et al., 2005). Under current manufacturing, use, and disposal practices this alternative is unlikely to increase the overall carbon store of the forest sector, which includes the forest and wood products derived from the forest (Harmon et al., 2009). Manufacturing, use, and disposal of harvested wood all entail significant carbon losses that are either as large as or larger than those in the forest itself (Krankina and Harmon, 2007). Wood products carbon offsets associated with biofuels and substitution of wood for more energy intensive building materials, such as steel and concrete, can theoretically increase the carbon “stores” of wood products beyond that stored in the forest itself (Perez-Garcia et al., 2005; Lippke et al. 2010). However, several issues need to be recognized regarding these offsets. First, most analyses have presented theoretical maximum product substitution offsets and ignored the effects of additionality (i.e., degree to which practices differ from business as usual or statutory requirements), permanence and replacement of existing wood products, and enduser preferences for building materials. If these factors are included, then substitution effects are substantially lower than the theoretical maximum and unlikely to surpass carbon stores in forests for many centuries if at all. Second, depending on the starting condition of the forest, both product substitution and forest-related biofuels can create carbon debts that delay carbon benefits. For example, biofuels harvested from existing forests could offset fossil fuel releases of carbon, but recent studies have indicated that carbon debts associated with the energy used during biofuel harvests, decreased carbon stores in forests, and differences in carbon to energy ratios could persist for decades to centuries, implying a significant temporal lag in net carbon uptake (Fargione et al., 2008; Searchinger et al., 2009). Third, being offsets, the effectiveness of both biofuel and product substitution will vary with the duration of the offset; the longer the delay in releasing fossil fuel carbon, the more effective offsets become: An offset with a 1 year delay would have little impact on atmospheric CO2 concentrations, whereas an offset of hundreds of years would have a much greater impact. Unfortunately, the duration of offsets is not well understood at this point, but it is unlikely to be infinite as tacitly assumed in many current analyses. Finally, while offsets are often counted as carbon stores, they are difficult to directly inventory because they are not physically in an identifiable location, whereas carbon stored in forests can be more directly inventoried and quantified.

    Sarah L. Shafer, Mark E. Harmon, Ronald P. Neilson, Rupert Seidl, Brad St. Clair, Andrew Yost 2011. Oregon Climate Assessment Report (OCAR) http://occri.net/ocar Chapter 5. The Potential Effects of Climate Change on Oregon’s Vegetation. http://occri.net/wp-content/uploads/2011/04/chapter5ocar.pdf

    I tend to give more credence to those who have done a good literature review.

    What is the importance of literature review?
    The importance of literature review include the following:
    i) use literature review to evaluate relevant articles’ research limitations & future research recommendation so that a graduate student can find his or her research topic.
    ii) use literature review to go through what are the theoretical frameworks in previous studies that can be ameliorated to form a new conceptual framework / research model.
    iii) use literature review to collect more information about their research topic before embarking on it i.e. no reinventing the wheel (in case it did, how to change the research topic / direction & again needs literature review),
    iv) use literature review to find the relevant research methods / instruments / survey questionnaire with high reliability & validity for adoption / adaptation.
    v) use literature review to provide supporting proof / precedent for an argument / claim during thesis examination / article review.
    vi) use literature review to justify an explanation based on empirical result outcomes when the graduate student is writing his or her thesis’s or article’s discussion section.
    In case you need to capture what literature you’d reviewed, you can take a look on a sample format of Literature Review Catalog through the following RG link:
    https://www.researchgate.net/publication/282604179_Excel_Template_for_Literature_Review_Catalog

    Reply
    • Well, then, it’s a good thing that we don’t base projects solely on incomplete carbon storage beliefs. I’m very content that the pluses outweigh the minuses in modern thinning projects in the Sierra Nevada National Forests. I wonder what the carbon numbers are like for the preservation of overstocked and unnatural forests, full of dead trees (like the Sierra Nevada).

      Reply
    • Thanks so much for engaging on this topic!!!! I really appreciate your taking the time to do this. Stay tuned and we will circle back around to taking yours and other papers and comparing them directly. Thanks again.

      Reply
  4. As far as national forests go, I think there is a simpler answer. Forest carbon is a “dominant ecological characteristic (for example, composition, structure, function, …).” It is therefore subject to the NFMA requirement of being managed within its natural range of variation. Since national forests must be managed for ecological integrity (which I assume would provide carbon within NRV), it seems like a fool’s errand to bother with trying to make a case for manipulating carbon as an independent purpose.

    Reply
  5. “Fire suppression/are campgrounds/ski areas are not “ecological characteristics.” They may occur if that can happen while NRV is met for dominant ecological characteristics. There should be an NRV for fire numbers, sizes and intensities; vegetation, fires (and humans) should be managed accordingly.

    Reply
    • Since we cannot significantly ‘manage’ humans, doesn’t that set of constants invalidate NRV? Or, at least base a new ‘small sample size’ NRV, with this set of constants in mind? Do we base projects more on current conditions….. or on NRV?

      Around here, in California, humans have manipulated the NRV since the glaciers retreated.

      Reply
    • Unfortunately, we are not able to manage fire with that degree of specificity (even if folks could agree on the details of what post- Ice Age NRV might have been, which time window, Native American actions included, etc.). If I recall correctly, NRV is only in the Handbook explaining how to define EI in the regulation.

      Reply
  6. The definition of IN the regulations of “ecological integrity” (which is the requirement for forest plan components):
    “Ecological integrity. The quality or
    condition of an ecosystem when its
    dominant ecological characteristics (for
    example, composition, structure,
    function, connectivity, and species
    composition and diversity) OCCUR WITHIN THE NATURAL RANGE OF VARIATION and can
    withstand and recover from most
    perturbations imposed by natural
    environmental dynamics or human
    influence.”

    The “right” answer is what a forest actually say it is. They just have to say it (and that must be based on the best available scientific information). (And there is no requirement to actually achieve the desired conditions, recognizing the lack of control of budgets, nature, people, etc.)

    Reply
    • Do we assume “Natural” to be a pre-human landscape? Or does it include historical human impacts and effects, too? If such an adjective mandates activity, should it have a rigid definition? I would go for ‘Desired Range of Variation’, based on science AND current conditions.

      Reply
      • “Do we assume “Natural” to be a pre-human landscape?”
        ===

        I’ve often found that the word/tern “Natural” means pre-white European introduction. Science has often viewed the Native Americans as a type of sub-human animal as part of the natural world’s ecosystem landscape where some sort of mystical balance was achieved.

        “Or does it include historical human impacts and effects, too?”
        ===

        No one has ever entertained the idea that Natives were equal to people today. If we remove the assumption they only used fire for conservation and stewardship of the land as part of the narrative, what impact did leaving a campfire unattended or improperly disposed of when moving on have on the landscape which may have sparked a wildfire ? Would it be considered natural because a native did it ? What impact did using fire by one tribe as a weapon against other Native Americans who were hated enemies have on the landscape ? Would it be considered natural because they were Native Americans ? What about irresponsibly starting a fire to cook food, send messages or to warm oneself during a Santa Ana or any other wind event have on the landscape ? Because it was done by the Native Americans, would their ignorance get a free pass and be labeled natural because they were Native Americans ?

        Reply
  7. The Planning Handbook (§23.11a) says, “The natural range of variation is a guide to understanding how to restore a resilient ecosystem with structural and functional properties that will enable it to persist into the future.” It really doesn’t lay out rules on any of this – the Forest Service just can’t be arbitrary; they need a reasonable explanation for what they decide NRV is. But they can’t ignore it or say “we’ll figure it out later.” It is a forest plan requirement by regulation.

    Reply
    • Then again, some people insist that ALL ecosystems are resilient and that humans should not be ‘meddling’. Some think that massive carbon releases are perfectly ‘natural’, and they should be welcomed. I think if we compared ‘persistent’ forest carbon (ie large fire resistant trees) from the 1400’s with today and it would be eye-opening that we are so very, very far from a coherent carbon policy in our western public forests. ‘Non-persistent’ carbon sits in trees that are imminently about to release it all, in the next inevitable human-caused incineration.

      Reply
  8. I agree that a national forest could decide to just let whatever happen, and the ecosystems would find a new sustainable trajectory. I haven’t yet seen this as an alternative considered in any forest plan revisions yet for some reason.

    Reply

Leave a Reply to Jon Haber Cancel reply