Arguments
"It is frustrating," says climate scientist Michael Mann from his office at Penn State University in the United States.
"There certainly has not been a hiatus in global warming — global warming hasn't stopped, even though you still hear those contrarian talking points," he says.
Professor Mann, the director of the university's Earth System Science Centre, is famous for his 'hockey stick' graph that reconstructed 1,000 years of global temperatures showing a dramatic spike towards the end of the 20th century.
The 'pause', also known as the 'slow down' or the 'hiatus', refers to the average rate of warming across the whole planet's surface in the last 15 years or so. The latest major report (pdf) from the United Nations Intergovernmental Panel on Climate Change said the rate of warming between 1998 and 2012 had been about 0.05°C per decade.
This rate, the report said, was "smaller than the rate calculated since 1951" which was 0.12°C per decade.
"The occurrence of the hiatus in global mean surface temperature trend during the past 15 years raises the two related questions of what has caused it and whether climate models are able to reproduce it," the report said (pdf).
This was proof enough for some commentators that computer models of the climate were wrong and that the risks of global warming may have been overblown.
Businessman and former ABC chairman Maurice Newman, the Prime Minister's top business advisor, has written that "temperatures have gone nowhere for 18 years".
Dr Scott Power, senior principal research scientist at the Bureau of Meteorology, was an author of the IPCC report.
"If you look at just these 15-year periods of globally averaged surface air temperature, then they fluctuate quite a bit," he says.
"We expect these fluctuations. What we know is happening is that the planet is warming in response to human increase in greenhouse gases very largely but there are fluctuations because of natural processes."
Power said that 1998, the start of the 'pause', was a particularly hot year due to the natural El Niño climate pattern that has a warming influence on worldwide temperatures.
Power said that if you choose a 15-year period starting in 1996 instead of 1998 then the rate of warming almost triples to 0.14°C per decade.
"Globally average surface temperature is just one measure of changes in the Earth's climate system," he says.
During the 15-year 'hiatus' period, studies of other aspects of the climate system have continued to show warming as expected.
The world's oceans have continued to gain heat, a recent study has found. And, in late March, a study in the prestigious journalScience found Antarctica's ice sheets were melting at an accelerating rate.
Many scientists have pointed to the recent extended periods of ocean cycles that are in phases that tend to have a cooling effect on temperatures at the surface of the planet.
The sun has also been in a state of unusually low activity, which can also have a cooling influence.
Volcanic eruptions generally cool the planet's surface. A study in the journal Nature Geoscience found that some climate models did not properly account for the higher levels of volcanic activity in the early 20th century. This meant that some models had overestimated the amount of atmospheric warming during the so-called slow-down.
Two studies led by scientists at the University of New South Wales found an increase in the strength of trade winds around the equator in the Pacific Ocean had pushed warmer surface waters deeper, having an overall cooling effect on global surface temperatures.
Dr Shayne McGregor, of the UNSW Climate Change Research Centre and an author of both the studies, said: "We have a certain amount of energy coming in at the top of the atmosphere and a certain amount of energy going out. With global warming we are changing that balance so the system is retaining heat.
"About 97 per cent of all the heat capacity of the Earth is in the ocean — that's where all the energy gets stored."
He said the increase in the trade winds was temporary. When they returned to normal, the rise in global temperatures would "come back faster".
A separate study by Dr Power found computer models were not simulating the changes in trade winds.
Professor Mann is a co-author of research published in February in Science that also found a role for the oceans in temporarily slowing the rate of global warming.
He said: "Our research reinforces work that tied the slow-down in surface warming — what I call the faux pause because it's not a real pause — with increased heat burial below the surface of the tropical Pacific Ocean.
"The problem is that it's temporary. We will fairly soon see the climate warm even faster than the models predicted."
An analysis by US government scientists published in Nature Climate Change suggested if temporary natural fluctuations were ignored then world was probably now warming at a rate of about 0.2°C per decade — higher than the IPCC's longer-term average. By 2020 this rate was expected to rise to 0.25°C per decade.
One group of scientists working at Australia's ARC Centre of Excellence for Climate System Science wrote last year that the term 'pause' was "ill-chosen and even misleading" in the context of climate change.
In the journal Nature Climate Change, the scientists showed during the 'hiatus' there had been an increase in the number of extreme hot days being experienced around the world.
"This is particularly relevant for climate change impacts, as changes in the warmest temperature extremes over land are of the most relevance to human health, agriculture, ecosystems and infrastructure," they wrote.
Click here to read the rest from Graham Readfearn at ABC Environment
Joel - Homewood has, not surprisingly, got his picture roughly backwards. From Santer et al 2014:
In short, that illustration shows ENSO and two large eruptions removed, but with remaining early 21st century volcanic forcings not included in the model forcings present. And that Homewood is quite wrong - there is an underlying warming trend over the last two decades, masked in part by volcanic variability that wasn't part of the model forcing sets.
In fact, you can see the error in Homewoods insistence on: "Ignore the misdirection about early twenty-first-century volcanic eruptions." In short, ignore the core of the paper so that he can mislead you. Nothing but denial.
Thanks, KR. Your comments have helped me to distinguish between what Santer et al. were trying to say and what Homewood wanted to say by carelessly (or intentionally) mis-interpreting Figure 1. Your comments prompted me to more carefully examine the paper by Santer et al. Although much of their paper involves statistical tests that are beyond me, my re-examination of their paper, combined with your comments, lead me to the following revised interpretation of Figure 1:
In Figure 1A (raw data), there's only a poor fit between CMIP-5 model predictions and the TLT data (RSS and UAH). In Figure 1b (ENSO removed), there's an excellent fit between data and model predictions up to 2002. That's because the CMIP-5 models were already adjusted to account for the El Chichón and Pinatubo eruptions but not for subsequent 21st-century minor eruptions. In Figure 1B, the model-predicted and satellite-observed cooling effects of the two big eruptions are evident in the big dips in 1983 and 1992. Notice that, when the effects of El Chichón and Pinatubo are also removed (Figure 1C), those big dips are remove, and the adjusted TLT data continue to agree with the adjusted model data until about 2002, but now both sets of adjusted data have consistently higher values (than in Figure 1B) from 1982 until 2000. In other words, removing volcanic influences causes the observed temperature data, as well as the model predictions, to increase during the time period of volcanic influence. One of the points that Santer et al. are trying to make is that failure to take cumulative 21st-century minor volcanic activity into account is what gives rise to the divergence between model-predicted (the models didn't account for these volcanoes) and observed (affected by these volcanoes) TLT data after 2000.
I think it's unfortunate that Santer et al. couldn't include a Figure 1D in which the effects of minor 21st century eruptions were removed from the TLT observations and then compared with the model predictions. I suspect such a figure would have shown significantly better agreement between models and data than in Figure 1C. In addition, such a figure would have shown a rising trend in the TLT data. I suspect that Santer et al. didn't want to create such a figure because our understanding and measurements of the effects of these minor eruptions are still too primitive to allow accurate graphing. Their doubts are reflected in the last sentence of their summary: "To reduce these uncertainties, better observations of eruption-specific properties of volcanic aerosols are needed, as well as improved representation of these eruption-specific properties in climate model simulations."
Thanks to your help, I no longer feel mystified by Figure 1, and I certainly no longer feel susceptible to Homewood's siren interpretation.
Joel_Huberman @52, two other minor but relevant points. First, the temperature series shown in Fig 1 C are only for 82.5 North to 70 South and therefore exclude part of the most rapidly warming region of Earth. Second, the observational records shown are tropospheric rather than surface records. The troposphere responds far more strongly to volcanic and ENSO influences than do surface records. Any dampening of the trend due to the small volcanoes in the satellite record will be greater than that at the surface. Ergo, it is likely that had surface records been used, and particularly truly global surface records such as GISS, or BEST, there would still have been a positive trend including the effects of the small volcanic erruptions.
What confuses me (and perhaps demonstrates something about eyeballing), is that from 1993, UAH and RSS seem pretty much in step, in the Santer graphic, but there's considerable disparity in the calculated trend.