Think before you sink
One major advance in the seemingly endless rounds of international climate negotiations was made by the Eighth Conference of the Parties (COP-8) meeting in New Delhi, India. The Clean Development Mechanism (CDM) was at last cleared for start-up as soon as enough countries ratify the Kyoto Protocol, which will probably be in early 2003.
But COP-8 also delayed, yet again, any final resolution of the formidable controversies and complications surrounding carbon sequestration or sinks due to land use and forestry projects. The right kinds of land/ forestry sink project have a huge potential under the CDM for improving the livelihoods of millions of rural communities, for protecting biodiversity, for bringing new life to degraded lands, as well as delivering large climate mitigation benefits. Yet this potential may never be fully realized under present CDM accounting rules for sinks, while final decisions on them have once more been postponed.
It was confirmed at COP-8 that draft decisions on one of the most critical sinks issues how to account for wood products will be prepared during the coming year for consideration by COP-9 in December 2003. That consideration, with no guarantees of final decisions, will occur more than halfway from the birth of the CDM in the 1977 Kyoto Protocol to the start of the first Kyoto commitment period in January 2008.
Your Bemused Observer has been looking into these matters and has uncovered an astonishing mess of peculiar assumptions, contradictions and misconceptions. These, amongst other things, underlie the adamant opposition by leading green non-governmental organizations to the kinds of sink project that could have huge potentials for multiple economic and environmental benefits: namely, productive tree plantations.
To start at the beginning, the idea of purging carbon from the atmosphere by growing trees was first proposed in 1977 by the physicist Freeman Dyson. In an article titled Can we control carbon dioxide in the atmosphere? he proposed that planting huge areas of forests as carbon banks could hold down atmospheric carbon dioxide for a few decades and buy time to complete a major shift from fossil fuels to zero-emission renewable energy sources. To do this, Dyson insisted, the forests must never be harvested (except to provide wood fuel as a replacement for fossil fuels) as this would rip open the bank walls and let the stored carbon escape.
A key feature of Dysons idea is that it considers only the forest site (and any wood products from the forest). Despite this limitation, the no-harvest forest has become the gold standard of the biotic-sink concept. The mature old-growth forest, rich in biodiversity, high in amenity value and protecting watersheds and soils, can store for modest investments a hundred tonnes or more of carbon, extracted from the atmosphere, on each hectare of land.
What virtues! And how much more virtuous it appears to be than its main rival under the CDM which limits land/forestry activities to afforestation and reforestation, that is, the young, biodiversity-poor, periodically-harvested, income-earning plantation. Under a wide range of conditions, the trees of a no-harvest forest appear to store roughly three times as much carbon as they would if they were regularly harvested. See Figure 1 above.
Furthermore, this superiority seems to be even greater when one includes the additional wood products resulting from periodic thinning and felling. A young plantation provides a variety of products, from construction materials and furniture to paper and firewood. In firewood-dependent developing countries in particular, many of these products are soon burned or dumped and allowed to decompose so that their contained carbon quickly returns to the atmosphere.
These ideas describe the received wisdom on this issue. However, as this article tries to show, it is more a case of received folly. That is, a sorry muddle of careless assumptions mostly about the accounting boundary one should use for land/forestry projects and the closely-related question of how one deals with wood products.
Unbelievably, land use/forestry scientists and analysts use at least three contrasting and incompatible approaches when trying to account for carbon stored or emitted by wood products.
1. All products are burned immediately after harvest
In this approach, you subtract all the carbon in harvested wood from the carbon sequestered by tree growth. This makes a sustainably-harvested plantation, like the one depicted in Figure 1, a zero carbon sink.
This unrealistic method is the default assumption in the present IPCC guidelines for reporting national greenhouse gas inventories. It is also advocated by some non-governmental organizations. For example, the August 2002 submission by Climate Action Network Europe (CAN-Europe) to the Climate Convention Secretariat on afforestation and reforestation under the CDM strongly opposes any attempt to award credits for carbon storage in wood products. This submission explicitly supports the IPCC guidelines.
2. Ignore all wood products
This approach is currently used for the CDM in which carbon accounting for afforestation and reforestation is restricted to carbon pools on the project site, which do not include off-site harvested wood. This approach is also used in scores of reports and articles which include those diagrams of saw-tooth carbon stocks as plantation trees grow and are felled (see Figure 1). One inevitable consequence is that, again, plantations come out as pathetic performers in the carbon-fixing stakes compared to no-harvest forests.
If site is defined with sufficiently wide geographical boundaries, or if carbon leakage outside the project site is estimated sensibly, there is hope that the CDM, using this approach, can properly reflect the real world and the large carbon-sink potential of most plantations. This crucial point is explained below.
There are, however, strong grounds to doubt that this will happen in the case of small-scale CDM sink projects of the kind that could bring huge benefits to rural people in the developing world. Attempts are being made to reduce the preparation costs of small CDM projects by simplifying their baseline and monitoring requirements. The simplifications at present include (1) limiting the project accounting boundary to the actual afforestation or reforestation site, and (2) ignoring all off-site carbon leakage.
3. Count all wood products
Plantations are credited with the carbon stored in all harvested wood products according to their lifetime within the wider economy. This requires estimates of how long each major class of products exists before it is burned or decays in a dump or landfill site, thus emitting carbon back to the atmosphere.
The method of calculation for this approach is widely used in two major carbon accounting models, CO2FIX (a sophisticated model intended to quantify carbon sequestration in forest ecosystems and wood products) and GORCAM (the Graz/Oak Ridge Carbon Accounting Model, developed under the Bioenergy Program of the International Energy Agency), as well as by some carbon sink projects and government institutions, one of which is the United Kingdoms Forestry Commission.
This approach is bizarre because it assumes that demand and consumption slavishly follow supply. That is, if a farmer cuts a plantation and sells, let us say, 100 tonnes of wood, people will respond by buying an extra 100 tonnes of timber, wood furniture, paper or whatever, compared to the no-harvest (no project) case. This is great news for book publishers, for example, who need only double their print order to double their sales.
The method makes some sense only in the extreme situation where wood shortages are so acute that one cannot buy wood or paper products until someone harvests another batch of trees. Each harvest will then result in more in-use wood products and hence a larger product (carbon) stock or greater emissions from burning firewood and so on. In most places, though, wood supplies are sufficiently ample that supply volumes track consumption, not vice versa.
However, this idea does bring us close to the real-world situation that carbon accounting methodologies must properly reflect although it does so for the wrong reason. The situation is essentially one of wood product supply having to match societys patterns of wood product consumption. This consumption pattern is not affected by small volumes of additional supply.
So when our farmer sells his or her 100 tonnes of plantation wood, because consumption is unaltered 100 tonnes of wood will not have to be cut from standing trees elsewhere. What these trees elsewhere would have supplied to meet consumption needs is instead supplied by the plantation. Compared to the case where there is no plantation, that is, there is no carbon-sink plantation project, these trees elsewhere will have an additional standing stock of 100 tonnes. In wood or carbon terms the stock loss from harvesting the plantation is balanced by a relative stock gain in the trees elsewhere. Total supply does not alter, because the total consumption pattern does not alter.
This perspective has dramatic implications for the comparative virtues of no-harvest forests and of harvest plantations.
First of all, if one extends the accounting boundary to include the trees elsewhere, the plantations are far more effective carbon sinks. See Figure 2, in which the carbon stock in the plantation plus off-site trees is nearly twice as great after eighty years as in the protected forest.
Second, in developing countries, elsewhere is very often natural forest. By diverting tree felling and wood removal from such forests, productive plantations help to conserve or enhance biodiversity. This is so, despite the assertions by Greenpeace, Friends of the Earth and other green non-governmental organizations that plantations are bad for biodiversity.
Third, carbon storage in plantations plus off-site trees is likely to be more permanent than in no-harvest forests. The likely impermanence of forest carbon sinks due to risks of fire, pest and disease is widely recognized as a major problem, both in realistic terms and in terms of procedures for allocating carbon credits to land/forestry projects. But these risks are much lower in plantations, which tend to be more carefully managed and watched over than protected forests. This is especially the case for the kinds of small-scale tree growing and agroforestry that can help to restore degraded lands, meet subsistence needs and earn income for millions of rural people.
Of course, reality is not quite as simple as this. A harvest of new wood supplies may lead to increased wood consumption: for example, when villagers can get ample firewood from a nearby firewood plantation rather than walking a long way to the forest and back. With large plantations, additional wood supplies may lower product prices and disturb demand and supply, perhaps over a large geographical area.
These effects might be difficult to measure and monitor, adding possibly unacceptable levels of uncertainty to CDM project accounting procedures. But this is not a good reason for ignoring what might often be a large carbon-sink (and biodiversity) benefit for CDM afforestation and reforestation projects.
The challenge is to find acceptable ways of including these benefits, perhaps through rules for estimating emission leakage from projects. Leakage the fact that project activities may result in altered greenhouse gas emissions outside the project accounting boundary is almost invariably thought of as a penalty that should be applied to a carbon sink project. The project inadvertently causes increased emissions or positive leakage. What we have seen here is that negative leakage, in large amounts, is probably likely in most afforestation and reforestation projects which also involve tree felling and harvesting; that is, in nearly all carbon-sink forestry projects other than those leave-them-alone Dyson forests.
Its an encouraging idea. Its also a daunting and possibly a bemusing idea for the people who must now draw up the wood product and leakage accounting rules for the Clean Development Mechanism.