On Thursday, July 27, 2017 at 12:35:05 PM UTC-4, Peter Nyikos wrote:
> On Wednesday, July 26, 2017 at 7:30:05 PM UTC-4, William Hyde wrote:
> >
> > I am not feeling at my best even now, so let me make a few brief points, after which you can ask any questions you like.
>
>
> I hope you don't mind my interspersing questions and comments below.
> I put them where I thought they best fit your narrative.
>
>
> > (I've actually co-written a speculative paper on global warming and possible distant future climate change:
> >
> > "Transient nature of late Pleistocene climate variability
> >
> > Thomas J. Crowley & William T. Hyde
> >
> > Nature 456, 226-230 (13 November 2008)"
>
> Wow, I'm impressed.
Well, I am retired. Fully retired. And this was my second last paper.
> > (1) Rate of change.
> >
> > In the 100,000 year ice age cycle global temperatures drop about four degrees C (some sources argue for 5 - note that I am talking about global average temperatures, not Vostok or Epica - quite a few graphs accessible via google say "global", but if you look at the source it is an ice core).
>
>
> What kind of data do you use to come to the conclusions below?
The estimated global temperature change between the last glacial maximum and today is mostly due to isotopic evidence from sea cores supplemented by other proxies such as ice cores (Vostok et al), speleothems, and so on.
The oceanic evidence (originally CLIMAP) is well distributed in space and it argues for about four degrees of cooling, some of the speleothem and other evidence argues for more. The ice cores give more, but that evidence is restricted to very high latitudes, a small fraction of the earth's surface so it is not heavily weighted.
Climate model simulations are not evidence, and given that interactive oceans were not a feature of models in the 1980s, it was no surprise that early model studies (1980s) gave about 3.5 degrees of cooling (Hansen et al) to about 4.2 (British met office, I think) as they used CLIMAP ocean data. It is reassuring, though, that more recent studies with far superior models give about the same result.
Our very simple model gave 2.2 as the effect of the high albedo of the ice sheets alone. Such a model is not able to account for changes in, e.g. CO2 or clouds (hence the low value) but it is independent of CLIMAP.
An ad hoc addition of CO2 change to the model gave, IIRC, 3.8 degrees.
All that said, if it turns out 25 years from now that the final verdict (to the extent that such a thing can exist) is six degrees I wouldn't be utterly shocked. But I suspect the result will be in the range of 3.5-5.5.
> > Let us take the larger number, and assume all the drop occurred in the first 10,000 years (the early cooling is the fastest). That is a rate of cooling of .05 degrees per century. If we take a very conservative view of global warming forecasts, the temperature between 2000 and 2100 will increase one degree.
>
>
> How reliable are these forecasts? What is the basis for them?
Utterly reliable, say some few, utter trash, say others.
Fairly reliable, say I and many others, including those who actually create the models.
Even the early versions of these models made a number of qualitatively accurate predictions. Manabe did the first such study that I am aware of, in the mid 1960s with a simple radiative-convective model. This model predicted that the lower atmosphere would warm, but also that the stratosphere would cool.
The first prediction, of course, has come true. But the second, which has also come true, is more interesting as it is counter intuitive. After all, if this is just a conspiracy, as our friend Eri maintains, why go out on such a limb? Why lie about something that is a priori unlikely to happen? (A review of upper atmosphere changes was published in Science, in somewhere around November 2006.)
And remember, these predictions were made 50 years ago.
Manabe and his colleagues moved on to more complicated models, early General Circulation Models. Without fully interactive oceans these models were distinctly handicapped, but they revealed another feature, polar amplification, which is evident today.
Models have also predicted that the warming would be greater in winter than summer, night than day. As observed.
So the models have predicted a number of features of the warming, some unexpected, they do a reasonable job of simulating past climates (not only the most recent ice age) and the warming we see today is well within the range of model predictions.
As the number I quoted above is only half of the ensemble model prediction for the 21st century, I'm rather sure we will warm at least that much.
> > Twenty times as fast. It is clear which is the immediate problem.
> >
> > Even if we take one quarter of a degree to be the likely warming in this century - consistent with some skeptical estimates - it is five times the ice age cooling. So it seems that an ice age would be easy to fend off, if we needed to.
>
>
> There was a theory back in the 1960's that the ice ages were brought on
> by increased snowfall due to the Arctic Ocean being open water and
> connected to the Atlantic between Greenland and Europe. We are making
> it open water a lot faster than it was in past ages, are we not?
That was the Ewing and Donn theory. I recall that the science classroom in my middle school had an article on this posted.
Ewing and Donn were respected scientists, to say the least (I've conversed with Donn, but not Ewing) but this was a strictly conceptual model. Conceptual models are highly valuable, and at one time they were all we had. But quantitatively this model does not work. Snowfall rates, ice flow, heat transport into the Arctic (an important part of the theory), none of it adds up to an ice age.
When I met Donn in 1982, he wasn't pushing it.
> Did you consider this now-forgotten (or so it seems) theory in making
> your estimates?
Well, for a start the temperature estimates I used for rates of cooling come from data, not theory. This conceptual theory, for that matter, would not provide any quantitative estimate I could use.
I don't know much about the history of mathematics. But is it not the case that interesting conjectures are made, but eventually shown to be false? If your field is like mine, even the disproof might lead to advances in the field.
Wally Broeker, one of the great geoscientists of the time, credits this theory with getting the scientific community more interested in ice ages:
"Donn would go around and give lectures that made everybody mad. But in getting mad, they'd really get into the subject" (Quote from memory and hence approximate, original in Weart's "The Discovery of Global Warming").
I see via a brief google that the Ewing and Donn theory is being promoted on skeptical websites, which is no surprise. But it has been dead scientifically for decades - though not forgotten.
There have been many theories of the ice ages (detailed in part in the book "Ice Ages - Solving the Mystery" by Imbrie and Imbrie which I recommended to Eri some months ago). Only the Milankovitch framework (the timing of ice ages is controlled by changes in the earth's orbit) has quantitative explanatory power.
>
> > (2) Uncertainty in timing.
> >
> > Statements like "Ice ages last 100,000 years, interglacials last 10,000" are at best approximations. It is not reasonable to use such approximate numbers to imply an ice age is coming *now*. An error of 1000 years is small compared to the actual uncertainty, but is a huge time in human history. A thousand years ago Cnut had just taken over England, Basil the Bulgar-slayer ruled in Constantinople.
>
> No argument here--this kind of prediction is extremely naive.
>
> > (3) Interglacials.
> >
> > Not only are all interglacials not 10,000 years, one at least extended to 30,000 years. Geochemist Warren Ruddiman
William Ruddiman, that is.
has an interesting if controversial theory as to why this interglacial is proving to be a long one (early human agriculture and elevated CH4 levels).
>
>
> What, didn't he make elevated CO2 levels part of his analysis?!?
Well, he's talking about greater warmth, compared to other interglacials, in pre-industrual times. But aside from CH4 he does consider CO2 N2O and other gases on which a farming economy can have an effect. But IIRC CH4 in this case dominates. Even if he is wrong, like Ewing and Donn he has opened up a new area for research.
> > (4) Orbital forcing. Ice ages are driven by variations in the earth's orbit. According to the late Dr Thomas Crowley, the best orbital configuration for glaciation that we are going to see for several millennia occurred about 800 years ago.
>
>
> Fascinating. That was about a century before the Little Ice Age began.
> Could the time lag be explained by various factors?
>
No, the orbital signal is too weak to produce so rapid a cooling as the little ice age.
The ice age and little ice age are very different events. If the Milankovitch theory is correct, the very early years of a new ice age see summer cooling, in mid and high latitudes but little cooling in winter, perhaps even warming, while the LIA was noted for very cold winters.
The causes of the LIA are far from resolved, but there was clearly a very significant volcanic component, the solar component was real (though to my mind often overestimated) and there were small changes in greenhouse gases - not an exclusive list. None of these effects are believed to be causal for a real ice age. (Though GHG changes are a positive feedback and an ice age already incipient might be hastened along by volcanic or solar change.)
>
> > In a sense we missed the on ramp for the next ice age, and must wait for the next one.
>
> "must" may be overdoing it.
Well, to extend the metaphor, we are still pretty close to the on ramp. If we did something a little illegal we might be able to get on.
Closely allied with that 1960's theory
> is a much more recent one: the Gulf Stream will wither and die
> without sufficient cold to make the waters sink where they are
> now sinking, in the vicinity of Iceland IIRC. This would bring
> much colder winters AND summers to Western Europe, retarding
> snow melt and thus increasing the earth's albedo in a positive
> feedback reaction.
Possible changes in the gulf stream are indeed a worry. For a long time even ocean modellers themselves felt that the tendency of the gulf stream to collapse in their models was a model failure. But this feature seems robust (note, I am not an oceanographer so this is second or third hand) and is being taken more seriously than it was in 1990.
Quantitatively it is not at all clear (to say the least) that it would lead to an ice age, particularly given the CO2 warming we are experiencing. If a weaker gulf stream cools Europe, that might not be a bad thing in a generally warmer world.
The Gulf stream does indeed change utterly in an ice age, flowing directly east at about 40 N. But despite this the ice sheets go much farther south in North America, and glaciation begins there.
>
> Whether this would be enough to bring on a genuine ice age, or
> just a repeat of the Little Ice age,
Or neither.
>I have no idea. Do you?
If anybody tells you exactly what will happen in such a case, they either know a lot more than I do or are lying to you. As I said above, the "little ice age" is not really a weaker version of the actual ice age, it is cooling for very different reasons. But if I could live long enough to collect, I would put a lot of money on weakness in the gulf stream *not* causing a new ice age or a new LIA, particularly when accompanied by GHG induced global warming.
We were all taught as children that "the gulf stream is the reason Europe is not as cold as Labrador". We were all taught wrong.
> > I'd like to comment on one of the quotes given. I was a professor at Dalhousie at the same time as Petr Chylek. He is a superb scientist and a natural skeptic (not just on scientific topics). He is a firm hater of anything smelling remotely of a bandwagon, and has bent over backwards to give a forum to GW skeptics, even organizing two conferences on the subject we are discussing.
> >
> > I have huge respect for him, and I hope that is clear.
> > But in this particular case his comment was at best beside the point. The report of increasing mass loss in Greenland in 2001 was sound. Greenland has continued to melt since 2001, 2002 was a year of record melting, and the melt has both continued and accelerated.
>
> I've seen a report that the snow and ice cover in Greenlad has increased
> this past year, especially in the southeast. Was that "fake news"?
Without looking anything up, I will assume this to be true because it fits a pattern. There is an implicit assumption that things are not getting worse unless every climate variable gets worse year over year. So any reversal is heralded as a sign that there is no problem after all.
But that is not the way the world works. There is a lot of "noise" in the system, so after a poor year the next may well be "better", but it does not mean the problem is going away or even getting less severe. It is what is expected - especially after record bad years like 2007 for arctic sea ice. The skeptics made a meal of it when 2008 was not so bad! But, as predicted, arctic sea ice extent has continued to shrink.
Now let me check. Yes, 2016 was a poor year (tenth worst on record), this year is showing less melt than any year since 2009. There's no reason to assume the trend is changing.
Surface melt does not of itself tell us how much mass the ice sheet is losing. More of it is worse, in general, but where and when the melt happens can be crucial, as some of the melt refreezes before it reaches the sea. And of course, the main mass loss actually occurs where the ice sheet meets the sea.
Total ice mass is measured by the GRACE experiment. Figure 3.4 here:
http://www.arctic.noaa.gov/Report-Card/Report-Card-2016/ArtMID/5022/ArticleID/277/Greenland-Ice-Sheet
shows how it has been changing.
Anthropogenic greenhouse gases are estimated to produce a warming over this century of .02C per year. This is an order of magnitude lower than the changes of shorter acting climate events such as ENSO, tropical vulcanism, and so on. It's not beyond the bounds of possibility that a combination of such events could produce a decade of distinct cooling.
Such a cooling would not be evidence that global warming is not happening (though of course it would be endlessly cited as such). Any more than evidence of a lower bank balance as you send your children to university implies that you've had a pay cut.
> Best wishes for improved health,
Thanks. It's just a persistent bug. I spent the winter disease-free so I must be punished in summer.
William Hyde