Extra climate targets urge faster CO2 cuts

University of Bern’s Marco Steinacher has helped show that setting limits on different aspects of damage from climate change will likely limit CO2 emissions more than just temperature alone. Credit: University of Bern

To give the world a chance of restricting damage caused by climate change, we need more than just a single temperature target, Swiss researchers have found. Marco Steinacher and his teammates at the University of Bern worked out the chances that climate change can be kept within harmful limits in six different areas. “Considering multiple targets reduces the allowable carbon emissions compared to temperature targets alone, and thus CO2 emissions have to be reduced more quickly and strongly,” Marco told me.

In December 2009, world leaders agreed the non-binding Copenhagen Accord, which ‘recognises’ that scientists think world temperature increases beyond 2°C above the pre-industrial average from 1850-1899 would be dangerous. It also mentions sea level rise, protecting ecosystems and food production. And as climate talks have continued since the 1990s, specific new dangers of CO2 emissions have been found. One serious impact that has been realised in the last decade comes from the fact that oceans absorb CO2 from the air, which makes the seas more acidic. That can make it harder for sea creatures’ shells to form, and together with warmer seas can damage coral, and in turn reduce fish numbers available for food. “Traditional climate targets have not addressed this effect,” Marco said.

It might seem reasonable to assume that negotiating climate deals on temperature limits alone could protect against other dangers. But until recently only very simple ‘Earth system’ models were available to test this against the idea of having several targets. They couldn’t simulate regional effects on quantities such as ocean acidification or farming productions, Marco said. “Climate targets that aim at limiting such regional changes can only be investigated with a model that has a certain amount of complexity,” he explained.

Keep it not-so-simple

To test the implications of having multiple climate targets takes a lot of simulations. “This is currently not possible with more complex models, so we have developed an ‘intermediate complexity model’ that balances complexity and computation costs,” Marco said. He underlined that their model has already been compared against measurements and other models to make sure it simulates the climate system well. “But of course, intermediate complexity models have their limitations and we encourage similar studies with more comprehensive models when this becomes computationally feasible,” the scientist added.

Marco and his teammates wanted to look at six areas that climate change is affecting, giving each four sets of damage limits, ranging from strong to weak. To do this, they needed to run many simulations because there is some uncertainty over  factors like the speed and efficiency of 19 processes included in their model. “For example climate sensitivity, the temperature response to a given increase in CO2 concentration, is not known exactly,” Marco said. “To capture this uncertainty we use 5,000 configurations that have different properties, such as lower or higher climate sensitivity, or slow or fast mixing of heat and carbon into the deep ocean.”

The rise in greenhouse gases caused by humans alters climate and ecosystems in a variety of ways, and the effects differ from one region to the next. Multiple climate targets are necessary in order to prevent dangerous interference with the climate system and thus negative social and economic effects. With the six climate targets proposed by the University of Bern climate researchers in their study, different environmental changes are limited that can have negative effects for humans and ecosystems on land and in the ocean. Credit: University of Bern

In a paper just published in leading research journal Nature, the researchers describe how they ran their simulation over the two centuries from 1800-2000 using these 5,000 different configurations. They then compared them against 26 different datasets of direct measurements of climate and carbon in the environment, and picked the most ‘realistic’ configurations. With the 1069 remaining simulations most accurately representing how the real world operates, the team ran them into the future under 55 different possible greenhouse gas emission scenarios. “We can then check in which of those simulations the targets are not exceeded and infer which emissions are compatible with the defined targets,” Marco said. The scientists use the range of outcomes across the emissions scenarios to give overall probabilities, for example from the ratio of scenarios exceeding the limits to those that don’t. And though they ended up running 65,000 simulations, their intermediate complexity model could do this in just a few weeks.

Revealing the big picture

The maximum quantity of CO2 that can still be emitted up until the end of this century through burning fossil fuels in order for the climate targets still to be met. In order to achieve all six climate targets jointly, the emissions have to be cut much more drastically than if global warming is only to be limited to 2°C. The grey part of the bars shows the amount of CO2 that has already been emitted in the past, while the red part shows the emissions still allowable up until the year 2100 in order for the corresponding targets still to be met. The red and grey part of the bar shows the expected emissions in this decade, assuming an annual rise in emissions of 1.8%. The uncertainties created by different assumptions regarding the future emission of substances other than CO2 are shown by the horizontal lines. Credit: University of Bern

The researchers found that to stay within all six limits in any of the four sets would mean stricter emissions limits than the temperature target, or any of the other targets, in the set alone. So for an emission scenario where global warming can stay within the 2°C limit in target set 2, there is a considerable risk that at least one of the other limits is exceeded. Ocean acidification limits put the tightest restraints on CO2 emissions, and strongly narrowed the  range of outcomes.

Joeri Rogelj at the Swiss Federal Institute of Technology, ETH Zurich, noted that the limitations of Marco’s team’s model mean its specific predictions must be used cautiously. But he points out the value of their results is in its general insights. “The study’s results clearly demonstrate the importance of holistic and integrated assessments of sustainable human development,” Joeri wrote in a “News and Views” article in the same Nature issue. “The conventional focus on temperature change alone should move towards a more comprehensive accounting of multiple objectives and their interactions, from the global to the local scale.”

To act on what that change of focus shows will likely need greater effort to avoid damage. “We have shown that including additional targets would probably lead to even more stringent emission reductions than reported here,” the Swiss team writes in its Nature paper. Marco also stressed that time is running out to make such effort. “Which risks we are prepared to take is ultimately a social and political questions but the constant rise in CO2 emissions is increasingly limit our options to act,” he said.

Marco and his teammates created an ‘ensemble’ of configurations from distributions of parameters for 19 factors they consider in their model. Then they compared the results against 26 observational data sets, picking 1069 of the most realistic simulations based on a score, S_m. After running these models under multiple greenhouse gas emission scenarios, they could work out probability distributions of allowable CO2 emissions for the defined targets. SAT is the surface air temperature target, SSLR is sea level rise. A_SO is a target for corrosiveness to shells of sea creatures, A_Ω>3 is a target for ocean acidity in coral reef habitats. C_NPP>10% is the fraction of the global cropland area that suffers from substantial local productivity reductions, and C_{carbon loss} is the percentage of carbon lost from cropland soils since 2005. Figure copyright Nature, used with permission from Steinacher et al, see reference below.

Journal references:

Marco Steinacher, Fortunat Joos & Thomas F. Stocker (2013). Allowable carbon emissions lowered by multiple climate targets Nature DOI: 10.1038/nature12269
Joeri Rogelj (2013). A holistic approach to climate targets Nature DOI: 10.1038/nature12406