From the UNIVERSITY OF COPENHAGEN – NIELS BOHR INSTITUTE
When we talk about climate change today, we have to look at what the climate was previously like in order to recognise the natural variations and to be able to distinguish them from the human-induced changes. Researchers from the Niels Bohr Institute have analysed the natural climate variations over the last 12,000 years, during which we have had a warm interglacial period and they have looked back 5 million years to see the major features of the Earth’s climate. The research shows that not only is the weather chaotic, but the Earth’s climate is chaotic and can be difficult to predict. The results are published in the scientific journal, Nature Communications.
The Earth’s climate system is characterised by complex interactions between the atmosphere, oceans, ice sheets, landmasses and the biosphere (parts of the world with plant and animal life). Astronomical factors also play a role in relation to the great changes like the shift between ice ages, which typically lasts about 100,000 years and interglacial periods, which typically last about 10-12,000 years.
Climate repeats as fractals
“You can look at the climate as fractals, that is, patterns or structures that repeat in smaller and smaller versions indefinitely. If you are talking about 100-year storms, are there then 100 years between them? – Or do you suddenly find that there are three such storms over a short timespan? If you are talking about very hot summers, do they happen every tenth year or every fifth year? How large are the normal variations? – We have now investigated this,” explains Peter Ditlevsen, Associate Professor of Climate Physics at the Niels Bohr Institute at the University of Copenhagen. The research was done in collaboration with Zhi-Gang Shao from South China University, Guangzhou in Kina.
The researchers studied: Temperature measurements over the last 150 years. Ice core data from Greenland from the interglacial period 12,000 years ago, for the ice age 120,000 years ago, ice core data from Antarctica, which goes back 800,000 years, as well as data from ocean sediment cores going back 5 million years.
“We only have about 150 years of direct measurements of temperature, so if, for example, we want to estimate how great of variations that can be expected over 100 years, we look at the temperature record for that period, but it cannot tell us what we can expect for the temperature record over 1000 years. But if we can determine the relationship between the variations in a given period, then we can make an estimate. These kinds of estimates are of great importance for safety assessments for structures and buildings that need to hold up well for a very long time, or for structures where severe weather could pose a security risk, such as drilling platforms or nuclear power plants. We have now studied this by analysing both direct and indirect measurements back in time,” explains Peter Ditlevsen.
The research shows that the natural variations over a given period of time depends on the length of this period in the very particular way that is characteristic for fractals. This knowledge tells us something about how big we should expect the 1000-year storm to be in relation to the 100-year storm and how big the 100-year storm is expected to be in relation to the 10-year storm. They have further discovered that there is a difference in the fractal behaviour in the ice age climate and in the current warm interglacial climate.
Abrupt climate fluctuations during the ice age
“We can see that the climate during an ice age has much greater fluctuations than the climate during an interglacial period. There has been speculation that the reason could be astronomical variations, but we can now rule this out as the large fluctuation during the ice age behave in the same ‘fractal’ way as the other natural fluctuations across the globe,” Peter Ditlevsen.
The astronomical factors that affect the Earth’s climate are that the other planets in the solar system pull on the Earth because of their gravity. This affects the Earth’s orbit around the sun, which varies from being almost circular to being more elliptical and this affects solar radiation on Earth. The gravity of the other planets also affects the Earth’s rotation on its axis. The Earth’s axis fluctuates between having a tilt of 22 degrees and 24 degrees and when the tilt is 24 degrees, there is a larger difference between summer and winter and this has an influence on the violent shifts in climate between ice ages and interglacial periods.
The abrupt climate changes during the ice age could be triggered by several mechanisms that have affected the powerful ocean current, the Gulf Stream, which transports warm water from the equator north to the Atlantic, where it is cooled and sinks down into the cold ocean water under the ice to the bottom and is pushed back to the south. This water pump can be put out of action or weakened by changes in the freshwater pressure, the ice sheet breaking up or shifting sea ice and this results in the increasing climatic variability.
Natural and human-induced climate changes
The climate during the warm interglacial periods is more stable than the climate of ice age climate.
“In fact, we see that the ice age climate is what we call ‘multifractal’, which is a characteristic that you see in very chaotic systems, while the interglacial climate is ‘monofractal’. This means that the ratio between the extremes in the climate over different time periods behaves like the ratio between the more normal ratios of different timescales,” explains Peter Ditlevsen
This new characteristic of the climate will make it easier for climate researchers to differentiate between natural and human-induced climate changes, because it can be expected that the human-induced climate changes will not behave in the same way as the natural fluctuations.
“The differences we find between the two climate states also suggest that if we shift the system too much, we could enter a different system, which could lead to greater fluctuations. We have to go very far back into the geological history of the Earth to find a climate that is as warm as what we are heading towards. Even though we do not know the climate variations in detail so far back, we know that there were abrupt climate shifts in the warm climate back then,” points out Peter Ditlevsen.
###
UPDATE: The article is Open Access at Nature Communications
http://www.nature.com/ncomms/2016/160316/ncomms10951/full/ncomms10951.html
h/t to Doug Huffman
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
An observation. If I want to prove man made global warming I start with the industrial revolution, circa early 1800s, and I ignore any temperature records from late 1700s, or any data about inter-glacial periods. Of course the temperature increases for awhile. It is all very natural. But if we are at the end of an inter-glacial, as has been suggested, the long term trend is down, not up.
Or am I missing something?
Arthur? Anyone?
Is my conclusion right?
Unknowable for another few thousand years, but sounds right.
Hmm. Thinking about climate as a chaotic system, re my long-winded previous posts, what would the state space of the system consist of? I can think of ocean and atmospheric temperatures, with the solar flux and cosmic radiation being system inputs. What else do you need to say “this is the climate”?
by using Takens theorem: take a time serie (x(t) of some proxy. You can reconstruct the phase plan by drawing x(t+tau) vs x(t). tau being the optimal time lag that you can define by several ad-hoc techniques (ie mutual information graph). As an example, this is what it gives when applied to the Vostok ice core data: https://dl.dropboxusercontent.com/u/56918808/Phase%20plan%20Vostok.docx.
Interesting! What’s the X axis? I couldn’t read it.
Thanks.
The Ox axis is the temperature at instant “t” while the Oy axis gives temperature at time “t + tau”. It is a visual way to check the “predictability window” of a system. For a purely deterministic system, the points fall on one well defined curve. For random data, they fillt he whole plan. For a dynamical (chaotic) system, the points are on trajectories around two or more “strange attractors” Finding the best value of “tau” is a little bit tricky. One way to do it is to find the first minimum in the mutual information graph (mutual information vs “distance” between points in the time series)
Thank you. Did you happen to read the Nicolis paper cited by Belousov below?
Thinking about it, this would explain the “snowball Earth” of some 650 years ago and the warm stretch for the dinosaurs. Something, probably continental drift, pushed the attractors far enough apart that switching between the two became very unlikely.
Nicolis and other colleagues from the Prigogine team were my professors of Physical Chemistry at the University of Brussels 😉
As I said earlier somewhere in this discussion, if you use a phase plan (built according to the Takens theorem) you can “visualize” the position of the two attractors and analyze the trajectories over time around them. Hurst exponents or Lyapounov exponents (more tricky to compute) are a way to describe these trajectories. The exact moment of switching from one attractor to another is driven by the nature of the system (no need for an external forcing) and is unpredictable. What can be predicted is the statisitical distribution of time intervals between switches. This has been found by a climatologist: the Edward Lorenz of the “butterfly effect”.
The author is making huge guesses and speculating with a very limited amount of research and understanding of these “fractal patterns” and earth’s climate systems. Not impressed.
For a good explanation of mathematical chaos, read “Jim”‘s site at http://www.chaotic-circuits.com/
especially http://www.chaotic-circuits.com/what-is-meant-by-chaos/
and http://www.chaotic-circuits.com/a-very-brief-history-of-the-discovery-of-chaos/
This important descriptive study by Ditlevsen gives useful new detail to the known fractality of climate on all timescales.
Comments that “fractals are deterministic” miss the point. Fractals are diagnostic of underlying nonlinear-chaotic dynamics.
In climate the most important implication of this is that climate changes BY ITSELF. The habit of looking for a discreet “forcing” for every wiggle of a climate curve is profoundly wrong, useless and misleading.
This was expressed better than I can in the following paper:
Nonlinear dynamic systems in the geosciences
G. Nicolis1 and C. Nicolis2
1Université Libre de Bruxelles
2Institut Royal Météorologique de Belgique
Abstract
Geophysical phenomena are often characterized by complex, random-looking deviations of the relevant variables from their average values. Typical examples of such aperiodicity are the intermittent succession of Quaternary glaciations as revealed by the oxygen isotope record of deep-sea cores of the last 106 years or the pronounced spatial disorder characterizing geologic materials. A major task of the geoscientist is to reconstitute from this type of record the principal mechanisms responsible for the observed behavior. Traditional approaches attribute the complexity encountered in the record of a natural variable to external uncontrollable factors and to poorly known parameters whose presence tends to blur fundamental underlying regularities. Here, we consider that complexity might be an intrinsic property generated by the nonlinear character of the system’s dynamics. We review bifurcations, chaos, and fractals, three important mechanisms leading to complex behavior in nonlinear dynamic systems, and stress the role of the theory of nonlinear dynamic systems as a major tool of interdisciplinary research in the geosciences. The general ideas are illustrated on the dynamics of Quaternary glaciations and the dynamics of tracer transport in a sediment.
Get the paper:
http://www.kgs.ku.edu/Publications/Bulletins/233/Nicolis/
‘Comments that “fractals are deterministic” miss the point. Fractals are diagnostic of underlying nonlinear-chaotic dynamics.’
It should also be noted that chaotic systems are also deterministic.
Of course, to complicate the matter there are external forcing inputs, which may or may not also be considered random, including changes in the Sun’s output, the particular galactic environment that we’re passing through at the moment, cosmic ray levels, and continental drift.
Very interesting paper, thank you.
Paul of Alexandria
It should also be noted that chaotic systems are also deterministic.
Yes but … when chaotic system are sensitive to initial conditions down to the quantum level then deterministic means not deterministic. It’s analogous to Pierre Laplace’s conjecture that if we knew the location and properties of all the particles in the universe, we could predict the future. What practical use is that?
Thanks for your endorsement of the paper, I haven’t read it fully yet, I’ll try to do so (maybe skipping the maths).
Ultimately, all chaotic systems are sensitive to that level; the question is: how far out can you go before the divergence between model and reality becomes too much? Weather is also chaotic, and we can go out a good week before things start to diverge too much.
While a good model of the climate won’t let us predict with accuracy too far, it would let us determine what the important characteristics are, where the attractors are, and what kind of excursion is most likely to result in an attractor shift.
Paul
Out of interest – is it known to what spatial scale chaotic systems are sensitive?
Does it go down to the atomic scale – or even smaller (e.g. quantum)?
Either in theory or experiment.
I really don’t know. I think that it depends on the particular system and the time scale involved. For instance, see https://en.wikipedia.org/wiki/Stability_of_the_Solar_System
“In 1989, Jacques Laskar of the Bureau des Longitudes in Paris published the results of his numerical integration of the Solar System over 200 million years. These were not the full equations of motion, but rather averaged equations along the lines of those used by Laplace. Laskar’s work showed that the Earth’s orbit (as well as the orbits of all the inner planets) is chaotic and that an error as small as 15 metres in measuring the position of the Earth today would make it impossible to predict where the Earth would be in its orbit in just over 100 million years’ time.”
Also http://hockeyschtick.blogspot.com/2013/09/chaos-theory-explains-weather-climate.html
Lorentz’s original investigations lead him to say that the molecular motion in the atmosphere keeps weather predictions to less than 3 weeks,
The various chaotic electrical circuits are effected by electrical and thermal noise, so I would think that they could definitely be sensitive to quantum effects.
It would be fascinating to figure it out.
Paul
Thanks for your informative answers.
The author says at the end we are entering a new period of warmth that might be unpredictable because we could be going from one regime to another. So far this hasn’t been proved and that is very important but it must be accepted that this is an accurate statement that unknowns to some extent are hugely endemic to this field. Is it not contradictory to say “it is settled” then? What is settled? Nothing.
It is not settled if this is unprcedented temperatures. It is not even known what is a temperature of the earth. I have several blogs on all these topics pointing out the many unknowns and many false predictions of this “science” which fails repeatedly to meet any predictions.
https://logiclogiclogic.wordpress.com/category/climate-change/