Sunday, January 24, 2016

Regional impact of global 2 Degrees C actually means far higher temperatures

Here is a good reason why we should be aspiring to meet the lower end of the Paris Agreement temperature goal of 1.5 degrees Celsius (2.7F).

In Paris diplomats adopted a temperature goal of "well below 2 degrees" of global warming and aspire to reach 1.5 degrees. This was a laudatory effort to bring climate policy into line with the climate science, which argues that beyond the 2 degree barrier we are likely to experience abrupt climate tipping points and enhanced climate feedbacks.

But 2 degrees Celsius average global temperature rise is a little misleading. "This climate target is abstract and invites misunderstanding,” says Sonia Seneviratne, Professor of Land-Climate Dynamics at ETH Zurich.

Seneviratne says that many people are likely to interpret two degrees globally as two degrees of warming in their region. They may then not push as hard for reducing CO2 emissions in their countries.

Seneviratne and her team at ETH Zurich in Switzerland worked with UK scientists, and Australian scientists from UNSW-based ARC Centre of Excellence for Climate System Science (ARCCSS). They calculated what an average global temperature rise of 2 degrees will mean on a regional basis from climate modelling.

Over 70 per cent of the earth's surface is ocean which spreads the average temperature, so this means for an average global temperature of 2 degrees Celsius, the average temperature over land mass and at higher latitudes will be much higher.

The scientists calculated the levels of extreme and average temperatures, as well as of heavy precipitation, that will occur in individual regions if the average global rise in temperature of 2 degrees Celsius is taken as a reference.

Graphic representations were done for each geographical region of the globe showing at a glance how average temperatures respond to the overall quantity of CO2 emitted and in relation to average global warming.

The scientists tested their new model using four examples: the Mediterranean, the USA, Brazil and the Arctic.

"We see starkly different rates of extreme warming over land even when global average temperatures reach just 1.5°C, which is the limit to the rate of warming agreed to at the Paris talks," said lead author Professor Seneviratne.

"At 1.5°C we would still see temperature extremes in the Arctic rise by 4.4°C and a 2.2°C warming of extremes around the Mediterranean basin."

For the Mediterranean example, the results reveal the following: if the global average temperature increases by 2°C, the region will see mean temperatures increase by 3.4°C on average. If, however, our aim is to limit warming in the Mediterranean to 2°C, then the global temperature must rise by no more than 1.4°C.

The highest regional temperature impact is in the Arctic where for global warming of 2°C, the average temperatures in the far north increased by 6°C. The researchers highlight that the 2°C target for the Arctic had already been exceeded when global warming reached the global average temperature increase of 0.6°C, probably around 2000.

In 2015 we appear likely to have reached 1°C of average global warming for the first time, helped by a strong El Nino.

The extreme regional warming projected for Alaska, Canada, Northern Europe, Russia and Greenland could have global impacts, accelerating the pace of sea-level rise and increasing the likelihood of methane releases prompted by the melting of ice and permafrost regions.

If emission rates don't change, areas around the Mediterranean, central Brazil and the contiguous United States could see 2°C of warming by 2030, more than a decade earlier than the widely accepted business as usual scenario, warn the scientists.

“The temperature difference between global average temperatures and regional temperature extremes over land not only has direct climate impacts, it also means we may have to reconsider the amount of carbon dioxide we can emit,” said co-author and Director of ARCCSS UNSW Professor Andy Pitman.

“For instance, to keep extreme temperature changes over the Mediterranean below a 2°C threshold, the cumulative emissions of CO2 would have to be restricted to 600 gigatonnes rather than the 850 gigatonnes currently estimated to keep global average temperatures increase below 2°C." warns Pittman.

In 2014 human generated Carbon Dioxide emissions totalled about 36 billion tonnes. "Carbon dioxide (CO2) emissions from fossil fuels and industry increased by 0.6% in 2014, with a total of 9.8±0.5 GtC (billion tonnes of carbon) (35.9 GtCO2) emitted to the atmosphere, 60% above 1990 emissions (the Kyoto Protocol reference year). Emissions are projected to decline by -0.6% in 2015 (range -1.6% to +0.5%)." explained the Global Carbon project.

The model projections also excluded unexpected changes in the climate system. It is very difficult to project when these abrupt changes may kick in. "What this research cannot take into account are abrupt climate shifts known colloquially as “tipping points”," said ARCCSS co-author UNSW's Dr Markus Donat.

"We have no way of knowing when our climate may change abruptly from one state to another meaning we could potentially see even greater regional variation than these findings show." warns Donat.


While most land-masses around the world will see an extreme temperature rise greater than 2°C, Australia appears to buck this trend. The modelled projections for Australia show little difference between global average temperatures and a change in its extreme regional temperatures.

"This might be something peculiar about Australia’s climate, or perhaps it highlights problems with the climate models," said Professor Pitman. "If the latter, there is a risk Australia will lack warnings about the increases in extremes that are now clearly available to Northern Hemisphere countries."

Most climate modelling has focussed on and been refined for the northern hemisphere. Pittman identified that this may be a potential hole in understanding of climate extremes in the southern hemisphere and needs urgent resolution with more focused model development.

So what are the current Mean temperature projections for Australia?

That depends on the emissions pathway (in UN parlance called Representative Concentration Pathways (RCP)) we adopt as part of the global Paris agreement to cut emissions. At the moment our targets make only a very small impact to the Business as usual RCP8.5 (purple) pathway.

In 2014 I used the CMIP5 model mean for RCP8.5 for 2050-2074 which gave results for Australia of a mean increase in temperature of 2.7C with precipitation showing no change. For the period 2071-2095 the CMIP5 mean model results showed a temperature increase of 3.8C and no change in Precipitation.

I decided to also run one specialised ocean-atmosphere model - GFDL-ESM2M. (See my reasons). Running the GFDL-ESM2M Single model using RCP8.5 for 2071-2095 resulted in an Australian mean temperature increase of 3.7C and a precipitation decrease 0f 0.4mm/day. So, slightly less temperature increase but a much drier climate with this climate model.

The CSIRO have done a Climate Futures tool to display climate variable projections for the different Australian regions.

Graphs show change in (from left) summer, autumn, winter and spring. Anomalies are given in °C relative to 1995(1986-2005) under RCP2.6 (Green), RCP4.5 (blue) and RCP8.5 (purple). Natural climate variability is represented by the grey bar.

RCP2.6 (Green) is the emissions pathway if we cut emissions strongly and rapidly. For political reasons we are unlikely at this stage to embark on this pathway, but it is the pathway we should be pursuing. We are currently proceeding at Business as usual which is approximated by RCP8.5 (purple). A moderate emissions pathway, RCP4.5 (blue), would be equivalent to adopting a target of 45 per cent emissions reduction by 2030. The graphs are part of the Summary Data Explorer

Southern Slopes Region - includes Melbourne and Hobart

East Coast Region - includes Sydney and Brisbane

Southern and southwestern Flatlands - includes Perth and Adelaide

Murray Basin

As you can see in all the temperature graphs the RCP4.5 emissions pathway makes a substantial difference in levelling out the temperature in the second half of the century while the Business as Usual pathway RCP8.5 the temperature just keeps climbing.

Most regional forecasts also detail reduced rainfall, with a notable difference between the intermediate and business as usual pathways. The Southern Slopes region is projected to reduce rainfall in spring months. Perth, Adelaide, Sydney and Murray Basin are likely to suffer rainfall deficits in winter and spring.

Here are the map projections for Mean Temperature changes to 2090. First the RCP4.5 emissions pathway, then the RCP8.5 emissions pathway map generated using the GFDL-ESM2M model. I chose using this model to be consistent with my results. If anything this model is a little conservative in comparison to the projections of some of the other models available to map using this tool.

Now which emissions pathway and climate targets do you think we should be pushing the government to adopt?