Study links natural climate oscillations in north Atlantic to Greenland ice sheet melt
Scientists have known for years that warming global climate is melting the Greenland Ice Sheet, the second largest ice sheet in the world. A new study from the Woods Hole Oceanographic Institution (WHOI), however, shows that the rate of melting might be temporarily increased or decreased by two existing climate patterns: the North Atlantic Oscillation (NAO), and the Atlantic Multidecadal Oscillation (AMO).
Both patterns can have a major impact on regional climate. The NAO, which is measured as the atmospheric pressure difference between the Azores and Iceland, can affect the position and strength of the westerly storm track. The study found that when the NAO stays in its negative phase (meaning that air pressure is high over Greenland) it can trigger extreme ice melt in Greenland during the summer season. Likewise, the AMO, which alters sea surface temperatures in the North Atlantic, can cause major melting events when it is in its warm phase, raising the temperature of the region as a whole.
If global climate change continues at its current rate, the Greenland ice sheet may eventually melt entirely–but whether it meets this fate sooner rather than later could be determined by these two oscillations, says Caroline Ummenhofer, a climate scientist at WHOI and co-author on the study. Depending on how the AMO and NAO interact, excess melting could happen two decades earlier than expected, or two decades later this century.
“We know the Greenland ice sheet is melting in part because of warming climate, but that’s not a linear process,” Ummenhofer said. “There are periods where it will accelerate, and periods where it won’t.”
Scientists like Ummenhofer see a pressing need to understand out how natural variability can play a role in speeding up or slowing down the melting process. “The consequences go beyond just the Greenland Ice Sheet–predicting climate on the scale of the next few decades will also be useful for resource management, city planners and other people who will need to adapt to those changes,” she added.
Actually forecasting environmental conditions on a decadal scale isn’t easy. The NAO can switch between positive and negative phases over the course of a few weeks, but the AMO can take more than 50 years to go through a full cycle. Since scientists first started tracking climate in the late 19th century, only a handful of AMO cycles have been recorded, making it extremely difficult to identify reliable patterns. To complicate things even more, the WHOI scientists needed to tease out how much of the melting effect is caused by human-related climate change, and how much can be attributed to the AMO and NAO.
To do so, the team relied on data from the Community Earth System Model’s Large Ensemble, a massive set of climate model simulations at the National Center for Atmospheric Research. From that starting point, the researchers looked at 40 different iterations of the model covering 180 years over the 20th and 21st century, with each one using slightly different starting conditions.
Although the simulations all included identical human factors, such as the rise of greenhouse gases over two centuries, they used different conditions at the start–a particularly cold winter, for example, or a powerful Atlantic storm season–that led to distinct variability in the results.The team could then compare those results to each other and statistically remove the effects caused by climate change, letting them isolate the effects of the AMO and NAO.
“Using a large ensemble of model output gave more statistical robustness to our findings,” said Lily Hahn, the paper’s lead author. “It provided many more data points than a single model run or observations alone. That’s very helpful when you’re trying to investigate something as complex as atmosphere-ocean-ice interactions.”
###
Looks as if this year’s summer minimum Arctic sea ice extent will finish between 2017 and 2015, higher than in 2016, 2012 and 2007, plus probably 2008. It might tie 2010 and 2011. The highest years of the past decade were 2009, 2013 and 2014.
http://nsidc.org/arcticseaicenews/charctic-interactive-sea-ice-graph/
This will be the sixth year in a row without setting a new record low. The three low years suffered cyclones in August. In 2016, two struck. The trend since 2012 is up, and flat since 2007.
Unless a new record low should occur from some freak WX event, it appears that the natural ~60 year (trough to trough or peak to peak) cycle bottomed out in 2012. The dedicated satellite record began in 1979, near the Arctic sea ice extent peak for the 20th century. The PDO flipped to its warm phase in the memorable WX year 1977.
“Using a large ensemble of model output gave more statistical robustness to our findings,” said Lily Hahn, the paper’s lead author. “It provided many more data points than a single model run or observations alone.”
There is something missing here (may not be missing from the article) which is that the range of the model outcomes is not mentioned. If the range is very broad, then the uncertainty about the “answer” provided is high. Very uncertain output do not attract much credence. There is no proof that simply having “a large ensemble” provides “robustness” if there is no robustness of the outcomes.
Further, to be robust, a model has to be validated. Running an unvalidated model thousands of times does not make the results or some average of them and getting the same answer thousands of times does not make it robust either. Robustness is showing that in varying circumstances, it provides the correct answer. Do that and I will be impressed.
As for observations, these are not given as much credence as they should, unless there is something inherently dubious about the methods or the observer. The sentence above gives the impression that lots of model runs are better than one set of good observations. The model is verified by those observations, not the other way round, or some other interpretation.
And the big error is that the model outputs are considered “data points” which is simple untrue. A model may be trained using data points, but it does not provide them until it has been validated. This is basic engineering protocol. One validates a model by showing that with two 9’s or three 9’s accuracy, it conforms to measurements (data) over the range in question. Maybe for climate we can accept one 9 because it is an immature field. But we don’t know if that happened without seeing the range of outputs generated.
Yes, it works the same way with experimental results too. One can often achieve a greater ‘precision’ by simply taking more (instrumental) data points with a higher sampling rate.
What does make sense in this report is the link between high pressure over Greenland and summer melting.
As high pressure over the northern Atlantic/Greenland has the effect of driving warm Mid-Atlantic air up along the western side of Greenland. Which during the summer can lead to melting.
There was a report in WUWT some months back about a study which lends support to this. Which that found that during the LIA that while europe cooled southern Greenland warmed. Which l think was largely linked to just this type of weather patterning.
I doubt Woods Hole will be allowed to continue with this line of research, as it would lead to the collapse of the global warming industry.
It’s easy to see why.
In this graph detrended HadCRUT lines up very well with the AMO. And the IPCC uses as their model validation century the years 1906-2005.
Look at where the cycle was in 1906. And look at where it was in 2005. That is an obvious artefact of about 0.4 C included in the ~0.74 C of “warming” last IPCC century (I’m citing the AR4 numbers). Remove more than half the warming as a natural and cyclic signal and the whole CO2 scam dies.
And that is even before the variance due to solar modulation of cloud cover is also considered.
Also the reverse of the Arctic sea ice extent cycle.
“Using a large ensemble of model output gave more statistical robustness to our findings,” said Lily Hahn, the paper’s lead author. Did this really pass peer review? If so, it tells you all you need to know about the peer review process.
It appears that they consider that natural processes are minor effects compared to Man’s contribution to climate change, rather than Man’s contribution being the parasite on the back of the flea sucking blood from the tail on the dog.
“The team could then compare those results to each other and statistically remove the effects caused by climate change, letting them isolate the effects of the AMO and NAO.“.
This process does not isolate the effects of the AMO and NAO. It isolates the models’ estimate of the AMO and NAO. IOW it tells them nothing about the climate, it tells them only what they coded into the models.
This amounts to an insurance derivative for woods hole as the multi decade decline of the AMO proceeds right under their slanted science noses.
Is the AMO weather or climate?
Jean;
Using the common 30-year baseline, a multidecadal oscillation would be climate if it’s over three decades.
“If global climate change continues at its current rate” we’ll be back to a glacial period while Trump is still POTUS
http://woodfortrees.org/graph/gistemp-dts/from:2017/plot/gistemp-dts/from:2017/trend
I find myself tuning out when something begins with ‘Scientists have known for years that warming global climate is melting the Greenland Ice Sheet, the second largest ice sheet in the world.’
While they may ‘know’ this, the aforementioned Scientists’ knowledge does not agree with the DMI SMB results since 1981, which they also know was at the end of a colder period.
DMI’s SMB charts show the Surface Mass Balance, which is the day-by-day modeled result of Snowfall accumulation, submlimation and melt, all of which go on all year-round. You can find the daily chart midway down the page at:
http://www.dmi.dk/en/groenland/maalinger/greenland-ice-sheet-surface-mass-budget/
The chart has 2 major displays, per the annotation below the chart: the ‘grey’ area represents the range daily high and low values for the years 1981-2010, a 30-year climatology, excepting the very highest and very lowest value for that day of any year. In other words, the span of the 2nd-highest to 2nd lowest values for those days of the year. The average for all the daily values, including the highest and lowest, is a darker grey line.
On top of that background, the day’s values are charted in Blue. For all the charts but 2011-2012, a line representing what happened in that year is presented for comparison.
Because the SMB report is of Accumulation, to know if the melting exceeded the snowfall for a year can be determined by looking at the last value for the year on the chart. The lowest year shown is 2011-2012, which had a net GAIN (new-sublimation-melt) almost 50 Gt of ice. All the other years since that one, for which charts are available, have higher amounts of gain, with the last 2 years gaining > 500 Gt. That’s right, the equivalent of 500 cu km of water.
As the DMI says, the SMB does not include the loss of ice due to calving of the glaciers. The 2016-2017 end-of-melt-season report offered the comment that the 550 Gt also offset the calving loss.
So, I offer the question: Do any of the DMI SMB results support the statement ‘Scientists have known for years that warming global climate is melting the Greenland Ice Sheet, the second largest ice sheet in the world.’
Well, it is the 2nd-largest ice sheet. But it is not melting more than the snow falls on it in any year that the DMI has reported. They must ‘know’ it from some other source…
“the Greenland ice sheet may eventually melt entirely” That’s when I stop reading. To think they got paid with taxpayers money and include a statement like that.
“Depending on how the AMO and NAO interact, excess melting could happen two decades earlier than expected, or two decades later this century.”
They should know that negative NAO/AO is directly associated with a warm AMO.
“We know the Greenland ice sheet is melting in part because of warming climate”
They don’t know that, and it doesn’t make sense as rising CO2 forcing should increase positive NAO/AO.
“Although the simulations all included identical human factors, such as the rise of greenhouse gases over two centuries, they used different conditions at the start–a particularly cold winter, for example, or a powerful Atlantic storm season–that led to distinct variability in the results.The team could then compare those results to each other and statistically remove the effects caused by climate change, letting them isolate the effects of the AMO and NAO.”
Sounds like they are irrationally removing only selected NAO noise and then attributing the residual natural AMO warming to AGW, without a shred of evidence that AGW is doing so. Any link to the paper?
Can someone correct me if I’m wrong, but I thought that recent instruments on satellites could do altimetery at the cm scale now, by comparing measured distances from relatively static points on Earth during their orbit to those that might be changing. I thought we know that parts of Greenland are melting while other are gaining ice. I am unwilling at this point to accept the premise that the Greenland ice sheet is melting on net until correctly measured data is in front of us.
Altitude measurements aren’t really good enough for that yet, and gravity measurements are strongly affected by GIA (i e that the ground beneath the ice is rising or sinking).
On balance Greenland has very likely been losing ice the last two decades, Antarctica is a different matter, nobody really knows whether it is gaining or losing ice. The uncertainty is much larger than the claimed gains/losses.
Greenland ice sheet may melt entirely but… Here is another forecast based on what is known as the Delphi method which is much more robust:
The G- ice wont melt entirely in any reasonable forecast time. The data I used is the fact we can still see ice that has been here for 2.7million years – the very beginning of the Pleistocene ice age and the Holocene period we are living in is comparatively cooler than previous interglacials.
https://www.sciencemag.org/news/2017/08/record-shattering-27-million-year-old-ice-core-reveals-start-ice-ages
Gary,
The previous interglacial, the Eemian, was much warmer and lasted thousands of years longer than the Holocene has done to date.
The Southern Dome of the GIS melted an estimated 25% more than it has so far in our interglacial, but the Northern Dome was virtually unaffected.
Don’t they talk to other scientists? On another post on this blog the Icelandic volcano is spitting out 20 kilotons of CO2 daily. What effect does this have on anything? More plants growing vigorously, I guess.