Earlier in the week, the Australian government submitted its provisional target (known as a ‘Quantified Emission Limitation or Reduction Objective’ (QELRO)) for the second commitment period of the Kyoto Protocol. The target, 99.5 per cent of 1990 levels, left many scratching their heads and wondering how it relates to the previously announced objective of reducing Australia’s emissions by 5 per cent below 2000 levels. What follows is an attempt at an explanation — warning, it contains technical content.
A QELRO places a cap on the cumulative net emissions from the relevant country over the commitment period. For Australia, the provisional 99.5 per cent cap relates to the period 2013-2020. This is obviously different from the 2020 target, which represents a single year objective.
In all previous government publications concerning the national caps for the period 2013-2020, the approach taken was to calculate Australia’s 2000 emissions, deduct 5 per cent to get the 2020 emission limit, and then assume a linear trajectory from 2012.
The same method was used in the latest Department of Climate Change and Energy Efficiency (DCCEE) projections, which were released earlier this month. There, the single year 5 per cent 2020 target was estimated to be 537 million tonnes (Mt) of carbon dioxide equivalent (CO2-e), and the 2013-2020 QELRO was estimated at 4482 MtCO2-e.
A quick comparison between the DCCEE projection and the Australian government’s submission shows up a discrepancy. The DCCEE QELRO is 4482 MtCO2-e; the Australian government’s submitted QELRO is 4626 MtCO2-e.
At first glance, this looks suspicious, making the critical mind wonder whether the government is up to its old tricks.
While there are grounds for questioning whether the government is doing enough, there are no dead bodies to be found in the QELRO submission — the reasons for the identified difference are technical not conspiratorial, and relate to three issues.
First, the DCCEE targets and QELROs were calculated using old ‘global warming potentials’ (GWPs), the metrics used to convert the different greenhouse gases into CO2-e. In the Australian government’s QELRO submission, it used the GWPs from the IPCC’s Fourth Assessment Report (see Table 1). The reason for the switch is that, in line with international reporting requirements, Australia intends to apply the new GWPs from 2015.
Table 1 Current versus new GWPs
As Table 1 shows, most of the new GWPs are significantly above the current ones, the two exceptions being nitrous oxide (N2O) and sulphur hexafluoride (SF6). All else being equal, the higher GWPs result in an increase in the QELRO.
The second reason for the difference is that, rather than assuming a linear trajectory from the 2012 emissions estimate to the 5 per cent 2020 target, the government drew a straight line from Australia’s average QELRO in the first commitment period, adjusted for the new GWPs.
Australia’s first commitment period (2008-2012) QELRO was 108 per cent of 1990 levels, or 2958 MtCO2-e, giving an average of 592 MtCO2-e per year. If you adjust for the new GWPs, and apply the 99.5 per cent target, you get a QELRO of 4540 MtCO2-e for the second commitment period, which is still significantly below the government’s estimate of 4626 MtCO2-e.
This leads to the third reason for the difference. In calculating the QELRO off the 1990 base year, the government has not used the estimate that was submitted for the purposes of the first commitment period. Rather, it has relied on its most recent estimate, contained in the 2012 National Greenhouse Gas Inventory.
The first commitment period estimate was 548 MtCO2-e, made up of 416 MtCO2-e from the energy, industry, agriculture and waste sectors and 132 MtCO2-e from deforestation. The revised estimate, using current GWPs, is 558 MtCO2-e; 418 MtCO2-e from energy, industry, agriculture and waste and 140 MtCO2-e from deforestation.
Those not accustom to accounting in the Land Use, Land-use Change and Forestry (LULUCF) sector might baulk at the size of the shift in the base year deforestation estimate (almost 9 MtCO2-e). Again, there is no conspiracy here.
LULUCF accounting is subject to a high degree of uncertainty. For deforestation, the estimation of emissions requires the government to identify all areas that satisfied the definition of forest on December 31, 1989 and detect which forest areas have been subject to permanent land-use change (i.e. a switch from a forest use (e.g. unmanaged lands, managed conservation areas, commercial forestry) to a non-forest use (e.g. grazing, cropping, infrastructure or urban) since that time. Having done that, it has to estimate changes in the live biomass, debris and soil carbon pools, and any related methane and nitrous oxide emissions, on the identified land units.
It doesn’t take Einstein to work out that even Einstein would have trouble doing this.
The government’s deforestation method uses satellite monitoring to track changes in land use and a series of models to estimate emissions. Due to gaps in the satellite coverage, the method has to randomly allocate detected clearing events across selected periods. As a result, every time the model runs, it produces a different result. Improvements in the processing of satellite imagery also lead to changes in the underlying activity data. The result is volatility (and significant uncertainty) in the deforestation emission estimates for the base year and all other years.
The combination of these three factors —new GWPs, a linear trajectory from the average annual first commitment period QELRO, and an adjusted 1990 base year estimate — allows for the reconciliation of the estimated second commitment period QELROs. In fact, if anything, the government has effectively increased its level of ambition by adopting the new approach. The direct conversion of the government’s submitted QELRO using current GWPs is 4445 MtCO2-e, which is approximately 37 MtCO2-e below the DCCEE estimate of 4482 MtCO2-e.
The final mystery in the whole QELRO affair is how the government is estimating the 5 per cent 2020 target. As stated, the most recent estimate using current GWPs is 537 MtCO2-e. This is 95 per cent of the 2000 estimate. But which 2000 estimate you ask?
As 2000 falls outside of the first commitment period, there is no official Kyoto Protocol emissions estimate for this year. There is a UNFCCC national total but the accounting rules for UNFCCC submissions are different, with most of the difference relating to LULUCF.
To get around this, the government has taken the most recent Kyoto-consistent estimate of emissions from the energy, industry, agriculture and waste sectors for 2000 and combined them with the UNFCCC-consistent estimate of emissions and removals from deforestation and reforestation for the same year (these were the only LULUCF activities that Australia accounted for in the first commitment period). Technically, the product is a ‘non-number’, having no official status and being consistent with neither Kyoto nor UNFCCC accounting rules.
If Kyoto accounting rules were applied for all relevant sectors, the 5 per cent 2020 target number is likely to be lower. While interesting, this is not of great significance because, in the end, the government picks the QELRO and how it’s calculated.
Andrew Macintosh is an Associate Professor at the ANU College of Law and the Associate Director of the ANU Centre for Climate Law & Policy.