RealClimate logo


Unforced Variations: Jan 2021

Filed under: — gavin @ 1 January 2021

According to the somewhat* arbitrary customs of our age, the 1st of January marks the beginning of a new year, a new decade and, by analogy, a new start in human affairs. So shall it be at RealClimate too**.

This month’s topics will no doubt include the summaries of the 2020 climate (due Jan 14th or so), ongoing efforts to understand and predict extreme weather in a climate context, and the shift by the weather organizations (WMO, NWS) to a new set of climate normals (i.e. moving from 1981-2010 to 1991-2020).

In the spirit of this new year, please make a renewed effort to stay vaguely on climate science topics, try to stay constructive even when you disagree, refrain from posting abuse, and don’t bother with cut-and-paste climate denial (that stuff was tedious enough when it was originally wrong, and is simply boring now). Thanks!

*completely

**Seriously, we are thinking about how to update/re-position this blog, and would welcome constructive suggestions from readers.

258 Responses to “Unforced Variations: Jan 2021”

  1. 251
    zebra says:

    jgnfld #248,

    This is the one I’ve been recommending; good interface for playing around.

    https://www.myphysicslab.com/pendulum/double-pendulum-en.html

    Referring back to my earlier comment and your and Ray’s replies, the predictability you describe for the pendulum and for climate both come from “the physics”. That’s why I think this is a good way to educate the “sincere unbiased seeker of truth about ACC”, if there are any such left.

  2. 252
    jgnfld says:

    @251

    Yes. I’ve seen that one too. To my mind, the chaos-makes-perfect-predictability-impossible-therefore-we-should-do-nothing argument is 100% wrong-headed at best, conscious disinformation at worst. I make no statement about the truth seeker. The climate just is not going to wander off to some random and completely unpredictable portion of the state-space simply due to chaos. With today’s data, theory, and methods, the uncertainty band has been slowly shrinking.

    For scientists, anyway.

  3. 253
    Piotr says:

    zebra: “the predictability you describe for the pendulum and for climate both come from “the physics”. That’s why I think this is a good way to educate the “sincere unbiased seeker of truth about ACC”?

    Weren’t you the one who argued that “climate system is chaotic”? You couldn’t mean that it has SOME chaos – because everything has “some” chaos – so it would be tautology. The only non-trivial interpretation of word “chaotic” is the one in which chaos DOMINATES. So: chaos dominates over physics, with the result being that the chaotic unpredictability dominates over physical predictability.

    So, are _you_ “sincere unbiased seeker of truth”? ;-)

  4. 254
    Piotr says:

    Mike(249):The study finds that sea surface temperature is the biggest driver of bleaching, while local efforts to improve water quality or restrict fishing have little impact on limiting its severity.

    severity yes, but there is some indication that they affect resilience (ability to recover after bleaching) – how quickly they can repopulate their zooxanthale (resident symbiotic algae) – the presence of other,/b> stressors can make it more difficult.
    Of course – if the heat waves become longer or more frequent this may affect the recovery as well.

    BTW, some people try find more high temp. resistant zooxanthelle species, less likely to be ejected by the corals (=bleaching). Which happens, if I am not mistaken – when the stressed out algae produce an irritant (radicals?) that makes the corals decide that they overstayed their welcome…

    Other try to edit temp. resistant zooxanthelle with CRISPR…

  5. 255
    Killian says:

    249 mike:
    29 Jan 2021 at 6:18 PM

    The future of the Great Barrier Reef – and other reefs around the world – will ultimately depend on how successfully we can limit ocean warming.

    This is the blunt conclusion of a new study, just published in Nature, which examines the impacts of recent coral bleaching events on Australia’s Great Barrier Reef. The event in 2016, for example, left just 9% of surveyed reefs untouched.

    The study finds that sea surface temperature is the biggest driver of bleaching, while local efforts to improve water quality or restrict fishing have little impact on limiting its severity.

    https://www.carbonbrief.org/coral-reef-survival-hinges-on-urgent-rapid-emissions-cuts

    We are quickly running out of time to prevent the disappearance of important global eco-assets, like coral reefs, to global warming.

    I will predict with no joy that the next major hot ENSO cycle will produce record-breaking loss of reefs.

    Mike

    Given all the stuff we all know about + increasing solar output + possible (very likely?) strong El Nino all coming at the same time, say mid-decade, sure, let’s stick to incremental non-systemic changes.

    After all, weather extremes and positive feedback loops in chaotic systems are predictable, manageable, and will likely lead to a peak in great new human achievements!

    Right…?

    I’m sure all will be fine…

  6. 256
    nigelj says:

    I thought the climate system was regarded as not a chaotic system (although with some chaos in it) while weather is regarded as chaotic, based on something I read that appeared well informed, but according to the IPPC “The climate system is a coupled non-linear chaotic system, and therefore the long-term prediction of future climate states is not possible. Rather the focus must be upon the prediction of the probability distribution of the system’s future possible states by the generation of ensembles of model solutions.” Presumably this is a “consensus” view?

    And how does the inability to predict long term climate state reconcile with the fairly well defined warming predictions by the end of this century? Or does long term climate state mean something over millenia level time scales?

  7. 257
    John Pollack says:

    Nigel @256 The IPCC is giving the consensus view. For example, in 2030, we can’t predict what state ENSO will be in, or any of the effects that it contributes to. We don’t know just how much arctic sea ice will be left at the end of the summer, or exactly where we’ll be in terms of ice loss from Greenland and Antarctica. That’s chaos.
    However, we can get a set of probabilities by running ensembles of climate models, and get a reasonable estimate. We can get some sort of estimate of the mean global surface temperature at the end of the century, but there is a widening spread of possibilities, as the pool of uncertainties grows.

  8. 258
    Piotr says:

    nigel(256):”Or does long term climate state mean something over millenia level time scales?

    I don’t think so – given that IPCC does not spend most of its time predicting effect of various emissions scenarios into millennia ahead.

    I think the problem is with the ambiguity in the term. Everything has some chaos – so unless we are quantitative, being “chaotic” becomes a tautology, hence meaningless.

    So we might limit what we call “chaotic” be quantifying what we mean:

    1. does chaotic unpredictability _dominates_ the physics predictability. Then even the ensemble modeling won’t help us much, as things become quickly intractably unpredictable. In this sense – weather would be chaotic, while the current global climate may not.

    2. by quantifying the effect of chaos in an analogy to std. dev.- here: departures of individual models runs from ensemble mean- so you could say that the local daily temperatures are MORE chaotic (bigger std. deviations of individual weather model runs) then climatic global temp.

    This would support your (and mine) intuition than global climate, the local weather is much more chaotic than global climate and answer to Monckton et al. who say: how can we predict global climate over decades when we can’t predict local weather after a few weeks (note to Zebra – _I_ am not saying that).

    3. by quantifying its relative importance compared to predictability – as an analogy to coefficient of variation (= std.dev/mean) in which you judge the relative importance of variations between individual runs of the models by comparing to mean value of the ensemble.