climate scepticism and ice cores

[Andre]

In the thread about mammoths and climate scepticism (http://www.physicsforums.com/showthread.php?t=333159), we have seen that the fossil remains of the Arctic Siberian mammoth steppe suggest that is was actually warmer than today, at least during summers. This mammth steppe existed throughout most of the final stages of ice age, the late Wisconian or Weichselian of the Pleistocene epoch, and raises question marks about the severeness of the cold of the ice age as inferred from the Greenland ice cores. Also an abundance of evidence was presented that the period between 17.5 and 14.5 thousand years ago, the Oldest Dryas or 'mystery interval', was also much warmer than expected. This is completely at odds with the paleo-temperature interpretations of the isotope ratios of the Greenland ice cores.

In this thread I'd like to discuss, (not by monologue if possible) why and how this could have happened by analysing the hydrographic processes that change the isotope signatures.

Things could get a bit technical, therefore I'd like to present some literature and textbooks first as reference and fundamentals of this little superficial analyses.

For isotope fractination processes I refer to these links/textboox mentioned in the sublinks.

http://wwwrcamnl.wr.usgs.gov/isoig/isopubs/

Furthermore for the analysis of the paleothermometer principles I like to use:
Jouzel, J. et al, 1997; Validity of the Temperature Reconstruction from Water Isotopes in Ice Cores; Journal of Geophysical Research Vol 102, No C12 pp 26,471-26,487, November 30 (http://www.ipsl.jussieu.fr/GLACIO/hoffmann/Texts/jouzelJGR1997.pdf),
one of the first publications that attempted to proof the robustness of the isotope paleo-thermometer.

more later.

[Andre]

Two different elements I'd like to give some more attention in this thread, seasonality of precipitation and general aridity. The first one is the simplest, so we start off with that.

Let's have a look at para 4.2 of Jouzel et al 1997 (http://www.ipsl.jussieu.fr/GLACIO/hoffmann/Texts/jouzelJGR1997.pdf), page 17, copied here for convenience.

http://i25.tinypic.com/2i6tuh.jpg

Obviously the isotope thermometer only works when it snows and every year is the weighted average of the mixture of the isotopic 'light' cold winter snow and isotopic 'heavy' warm summer snow (see also fig 1 of Jouzel et al). If the ratio in summer and winter snow changes then so does the average isotope signature. For instance if the warm summer snows diminishes in an arid period, only the cold winter snow remains, giving a cold bias. Jouzel et al acknowledge that real time observations confirm this but revert to models to conclude that there are no reasons to assume such a shift for the ice ages. So models right, reality wrong?

However, their models were as good as the knowledge of those days, when Siberia was a blank spot, assumed to be covered by a lot of ice most likely. However, the modern fieldwork in Siberia has generated quite a different picture, which should put some question marks to the accuracy of these assumptions and consequently the models.

So it seems that seasonalitly of precipitation could have played a role after all.

Alother element what Jouzel et al do not address is the effect of general aridity. This is more complex and I'd like to address that tomorrow.

[sylas]

I'm not likely to be a major participant in these threads; but there are a couple of aspects of the discussion that are a trifle... odd.

My own main focus will probably continue to be basic physics as it relates to understanding of climate in the present, where we have much better measurements and understanding of the processes involved; and also where there is a combination of widespread public interest along with a large amount of distracting misinformation founded on profound misunderstandings of quite straightforward underlying physics. I'm not a really driven by an interest in climate specifically, so much as an interest in fundamental physics and public science education.

But I've looked a bit at some paleoclimate stuff; and I want to throw out some general thoughts.

Who makes a good skeptic?

In my experience, the best skeptics are working scientists actively involved in the topics we are examining. Individuals can develop a blind spot of course; but in a field like climate and paleoclimate there are all kinds of competing ideas and active research questions, so using the scientific literature is a very good way to get at the open and active questions.

Can we do that in this thread also? Of course. There is a substantial literature on oxygen isotope temperatures. The 1997 reference given here is a very good starting point for the issues; and if you don't know off hand what developments have taken place since then in testing and calibrating the oxygen isotope thermometer, a citation search for texts referring to this paper would be a way to look.

Here are three possibilities:

Schlosser E., et. al. (2008) "Atmospheric influence on the deuterium excess signal in polar firn: implications for ice-core interpretation", in "Journal of Glaciology" Vol 54, Iss 184, pp 117-124. (pdf (http://epic.awi.de/Publications/Sch2008a.pdf))
This notes some of the difficulties with the oxygen isotope thermometer that Andre has mentioned, in fairly general terms.
Schlosser, E, and Oerter, H. (2002) "Seasonal variations of accumulation and the isotope record in ice cores: a study with surface snow samples and firn cores from Neumayer station, Antarctica", in Annals of Glaciology, Vol 35, pp 97-101. (pdf (http://imgi.uibk.ac.at/IceClim/IceClim/Polar/schlosser-etal2002b.pdf)). A specific quantified study.
Masson-Delmotte et al. (2008) "A review of Antarctic surface snow isotopic composition: Observations, atmospheric circulation, and isotopic modeling", in J. Climate, Vol 21, pp 3359-3387, doi:10.1175/2007JCLI2139.1. (Abstract and download at NASA (http://pubs.giss.nasa.gov/abstracts/2008/MassonDelmotte_etal.html)). Very detailed. Both Jouzel and Schlosser are also listed as co-authors.

I have not read these papers in detail, but a quick scan suggests that they may be directly relevant to the topic, of sorting out potential errors and biases in the oxygen isotope thermometer.

It's a good thing to have posts that can explain the issues more simply than you might find in papers, but it's still a requirement that the issues and criticisms raised can be supported from the scientific literature.

It should be an honourable thing, in general, to be a skeptic. But in climate the word has become somewhat debased as it has often been turned on its head and applied to extreme credulous naivety which will seize upon any criticism which can somehow be spun into a basis for avoiding the conclusion that human activity is at present driving a shift in global climate towards generally warmer global conditions. I tend to discriminate between a real skeptic and a faux-skeptic largely on how well they discriminate between credible and nonsensical objections to the ideas they are skeptical about.

To help avoid some of the frivolous parts of the discussion that take place, physicsforums has in place a requirement that we use peer reviewed literature as a basis for argument. It doesn't mean pulling out a paper and saying what is wrong with it. It means that if you want to dispute the conclusions of a paper, then you should be able to show that the criticisms are not just your own ideas, but are themselves part of the scientific discourse. That hasn't happened in this thread yet. There's no reason it can't happen -- unless of course the objection is that everything about paleoclimate is wrong or unclear and therefore anthropogenic warming is unproved. In which case we are off the deep end and into nonsense again.

I have no official standing in physicsforums or climate science, but hopefully these links can help bring things into line with what I understand as the requirement. It's a good topic; I'll read even if I don't post any more.

Cheers -- sylas

[Andre]

Before addressing studies about the actual observed variation of isotopes in the ice cores, we need to have an idea of the processes in the water cycle, that brings snow to the ice sheets.

Again the physics of the fractination of 'water' isotopes during transitions are illustrated here (http://wwwrcamnl.wr.usgs.gov/isoig/res/funda.html) and here (http://wwwrcamnl.wr.usgs.gov/isoig/isopubs/itchch2.html) and more specifically here (http://wwwrcamnl.wr.usgs.gov/isoig/isopubs/itchch2.html#2.3.1),

quoting

During phase changes, the ratio of heavy to light isotopes in the molecules in the two phases changes. For example, as water vapor condenses in rain clouds (a process typically viewed as an equilibrium process), the heavier water isotopes (18O and 2H) become enriched in the liquid phase while the lighter isotopes (16O and 1H) remain in the vapor phase. In general, the higher the temperature, the less the difference between the equilibrium isotopic compositions of any two species (because the differences in ZPE between the species become smaller).

So let's follow an air mass generating water vapor from the ocean. The first process we encounter is the fractionation of the water molecules during the evaporating, interestingly enough we see: (http://wwwrcamnl.wr.usgs.gov/isoig/isopubs/itchfig2-5.html)

http://gallery.myff.org/gallery/121696/Fig2-5.jpg (http://gallery.myff.org/gallery/121696/Fig2-5.jpg)

Evaporation from an open-water surface fractionates the isotopes of hydrogen and oxygen in a manner which depends on a number of environmental parameters, the most important of which is the ambient humidity.

Evidently, the more arid the stronger the fractionation, hence in low humidity conditions the water vapor is already more in heavy isotopes than under moist conditions.

Then the airmass transports the water vapor towards the ice sheets, normally from warmer lower lattitudes to the cold arctic cores. hence the air cools and condensation takes place, causing fractination again, the heavy isotopes condense first; depending on temperature, the colder the stronger the fractination.

Furthermore as the air mass keeps cooling progressively on its way to the poles, more and more 'heavy' water 'rains out' leaving the lighter water vapor behind. So the precipitation becomes increasing 'lighter' as the heavy molecules disappear. This is known as the Raleigh effect (http://wwwrcamnl.wr.usgs.gov/isoig/isopubs/itchch2.html#2.3.3).

So as these processes are temperature dependent, we see that the temperature of the source water, ambient atmosphere and clouds all play a role in the rate of fractination in this process.

to be continued.

[Andre]

So what would happen if we look at the same process but we decrease the absolute humidity of the source area of the water vapor? This can be due to several processes, variation in (sea) surface temperature (lower) as well as lack of wind and stability of the atmosphere (inversions) decreasing the rate of mixing with higher layers. A simple calculator here (http://rpaulsingh.com/problems/Ex10_2.htm).

So this air rises by convection or advection (http://www.ace.mmu.ac.uk/eae/weather/older/Convection.html) until the saturation point at which clouds start to form. Now the fractionation of the heavy isotopes compensating depends on the local temperature at that particular moment. Do we know that temperature, given a known initial absolute humidity (http://www.engineeringtoolbox.com/absolute-humidity-air-d_681.html)? This may be an interesting question as we will see later.

In other threads we have been talking about dry and moist lapse rates and condensation. The initial temperature at which condensation (http://www.ace.mmu.ac.uk/eae/weather/older/Condensation.html) takes places is known as dew point (http://www.ace.mmu.ac.uk/eae/weather/older/Dew_Point.html).

But the dew point is mostly dependent of the absolute humidity of the air in the initial conditions, and drier air has to cool more to a lower absolute temperature before condensation takes place. Therefore the isotope ratio of the condensed water is basically a proxy for the dew point which is a result of the absolute humidity of the source air.

So next, this parcel of air travels to the poles to deposit its final snow on the ice sheets. So during this transport the Rayleigh effect causes the heavy isotopes to rain out first, leaving the remaining water vapor depleted with heavy isotopes and the vapor get lighter, but since this happens at lower temperatures higher in the atmosphere, the depletion is quicker than under the normal conditions and finally isotopic much lighter (and less) snow accumulates on the ice sheets. All effect of initial drier conditions.


Notice the ambient surface temperature are not mentioned in these dynamics, only the dew point as function of absolute humidity, causing the condensation taking place in higher - colder layers of the atmosphere. Of course the ambient surface temperature does play a role in forming the initial absolute humidity, but so do other factors, and the Sahara dwellers would probably not agree if one would state that humidity is correlated directly with temperature.

So next it would be interesting to see if the early studies about the isotopes in the ice cores acknowledge the function of the absolute humidity of the source in the isotope dynamics.

[Andre]

So, as demonstrated, apart from the Rayleigh effect, the main determining factor of isotope fractionation in precipitation is the temperature of first condensation, which is directly depending on the absolute humidity. So we take a good look at Jouzel et al (http://www.ipsl.jussieu.fr/GLACIO/hoffmann/Texts/jouzelJGR1997.pdf) to see if they acknowledge this dependence.

In the attempt to quantify isotope fractionation one would expect elaboration upon this. However in para 3.1. page 26,478 (7) we find:

http://gallery.myff.org/gallery/596225/jouz-1.jpg

They identify source temperature and condensation temperature, apparently as independent variables, whereas I have indicated that there is a relation between those two, since source temperature is the more prominent variable for the absolute moisture of the air above it, which in turn determines the ‘temperature of condensation’ or rather “dew point”.
Also the discussion about cloud-surface temperature difference, para 4.4 pag 26,483 (13), is about two elements,

First, it is the temperature during the precipitation event that is imprinted in the isotopic signal. Second, the formation of an inversion layer of cold air up to several hundred meters thick over the polar ice sheets () makes the temperature of formation of precipitation warmer than the temperature at the surface of the ice sheet.

But it’s not about the temperature difference in the source area due to differences in absolute humidity.

But things get more compelling if we have a look at fig 4 (para 3.1.), where they compare the isotope signature with the snow accumulation rate:

http://i25.tinypic.com/2r7xdsw.jpg

The extreme high correlation is explained thus:

http://i32.tinypic.com/3312l9l.jpg

However the most direct cause for variation in accumulation is not addressed, as hinted before, changes in humidity regime. So it appears that fig4 supports the alternate idea here that absolute humidity at the moisture source could be (one of) the major contributor to the isotope signature in the ice cores.

So if that was to be true, we should be able to test that, we have seen already in the other thread that the Northern Hemisphere temperatures in the timeframe 14.5 – 17.5 thousand years ago, were up, while the Greenland isotopes as seen here, were down, challenging the isotope thermometer.

How about that second low isotope period roughly between 12.7 and 11.6 thousand years ago?