Tuesday, January 23, 2007
Cooking the greenhouse books
THE GOVERNMENT’S OFFICIAL opposition - the Australian Institute (AI) - has again challenged the Howard administration, this time accusing it of cooking the books greenhouse gas emissions.

The AI has highlighted significant discrepancies between land clearing rates as calculated by the [federal] Australian Greenhouse Office (AGO) and the Queensland government’s Department of Natural Resources and Mines, with the federal office consistently estimating lower clearing rates when compared to State figures.

This has huge bearing on whether Australia can meets its ‘commitments’ under the Kyoto Protocol.

Land Use, Land Use Change and Forestry (LULUCF) was belatedly included in Kyoto carbon accounting mechanisms, on the insistence of Australia, whose delegation threatened to walk out if their demands were not met.

The ‘Australia Clause’ – as it came to be known - not only allowed land use changes to be included in carbon accounting, but also for the net emissions from LULUCF in 1990 to be included in the baseline, against which total emissions would be measured in the commitment period.

The clause did not significantly affect most developed (Annex 1) nations, as they generally were undergoing a long period of reforestation following almost total clearing since the Industrial Revolution. Of Annex 1 countries, only Australia continued to clear native forest at significant rates and benefiting from ‘avoided deforestation’ was going to be a huge boon.

Reducing clearing could effectively offset growth in emissions from other sectors as a result of population growth and GDP expansion.

When the Kyoto Protocol was agreed in 1997, the Australian government was well aware that its LULUCF emissions were already falling fast from a very high 1990 baseline. Coupled with the 8% allowed increase in emissions negotiated by Australia, the government effectively had to do very little else to meet Kyoto objectives.

But here comes the rub. (And the acronyms.)

To calculate clearing rates, AGO uses a complex process involving satellite imagery and computer and human checking of images known as the National Carbon Accounting System (NCAS). This system measures tree cover against Intergovernmental Panel on Climate Change-defined standards for reforestation, afforestation and forestry. It has been peer-reviewed and meets minimum standards for methodology as set by the IPCC.

The Queensland government publishes its land use data in the Statewide Landcover and Trees Study (SLATS). Again, this involves a highly complex methodology, which interestingly also deploys ‘ground-truthing’ to verify tree cover changes. Importantly, it uses subtly different definitions of tree cover to that defined by Kyoto and employed by NCAS.

However, despite the two data series measuring approximately the same thing (tree cover), there are huge differences in the two data series, with the federally estimated clearing rates consistently and significantly below SLATS figures. Furthermore, while NCAS clearing rates are trending downwards, SLATS shows an upwards trend.

True, the methodologies remain slightly different, but surely tree cover is tree cover is tree cover? Either it is there, storing carbon, or it is not, as it has been pushed over to make way for another privatopia or more pasture.

Though NCAS system does not account for some kinds of regrowth (if not human promulgated), AI argues that this does not account for the high level of discrepancy. (Kyoto carbon accounting considers only anthropogenic changes in land use, not actual changes on the ground, as measured by SLATS.)

The government is backing its figures. The AI is calling for an independent review of NCAS, questioning its validity.

BUT THIS IS not the only issue AI has with NCAS. In 2000, the National Greenhouse Gas Inventory estimated the 1990 baseline for net carbon emissions from LULUCF to be 113Mt CO2-e. By 2004 (the latest figures available) this had been revised upwards to 129Mt.

Out of nowhere, the Howard government has ‘found’ another 16Mt of CO2 emissions per year to play with. Alas, this has been quickly gobbled up by runaway emissions in all other accounting sectors, particularly stationary power generation.

The “2006 Tracking to the Kyoto Target” report declares Australia is no longer on course to meet its commitments. In 2010, CO2 emissions will be at 109.25% of 1990 levels, more than 1% over the target. (Of course, Australia never actually signed the Protocol’s ambitions into law, only ‘committed’ to meeting them in a relaxed, non-binding sort of way.)

But using the figures for LULUCF calculated in 2000 (113Mt), by 2010, Australian emissions would be at 112.5% of 1990 levels. It seems NCAS’s continued downwards revision of the LULUCF 1990 baseline is providing considerable breathing (or polluting) space for Australia’s greenhouse gas emissions (see below).



To refute NCAS is a considerable challenge and will require the kinds of resources and expertise that is only employed by the AGO. Instead, maybe we should just believe what the Howard government tells us about rates of land clearing and carbon storage.

However, it all seems a little too convenient to be accepted without question, especially in the context of Howard’s belligerent anti-Kyoto position.

Each year the net LULUCF baseline gets revised down. Each year the projected emissions from the energy sector go up. For example, in the 2005 annual report, even with a lower LULUCF 1990 baseline, Australia was on track to meet its 108% target by 2010. This year, even with a higher LULUCF baseline, Australia is going to miss its target, thanks to a blow-out in industrial and energy emissions.

To expand on a cliched analogy: Not only is the government moving the goal posts to its advantage, at the same time it is making the goal bigger.

But sadly, it is still missing.

---