dumb question on ice cores

[Indy Lurker]

It seems like the most recent ice cores show CO2 levels of 280 ppm, when the current air samples show 380 ppm.


For consistency, can't they / shouldn't they take some recently made ice and test it so they have current samples using the same method as all the other samples?

[madokie]

1. no answers but a question for you, when was that last sample taken?

[wtmusic]

2. That's a good question

I'd have thought they would have taken some as controls too. Where are you getting your data from?

[RaleighNCDUer]

3. Wouldn't that "280 ppm" depend on when that ice was laid down?

I imagine that 1000 yrs ago (medieval warm period), 10,000 year ago (end of the ice age), and 100,000 years ago (deep in the ice age) all had significantly different readings.

[jpak]

4. Glacial ice takes time to form

There's a progression from new snow -> firn (old snow pack) -> ice that has to occur before atmospheric gases are trapped in ice bubbles.

The time it take for this to occur varies depending on location - years to decades.

The Mauna Loa Times Series is the best record of recent changes in atmospheric CO2 and it accurately blends into the ice core record.

[hatrack]

5. If they're really the most recent cores, they're not showing 280

Ice cores going back a few decades match the Keeling Curve, which is the best contemporary data source for atmospheric trace gases.

[Nihil]

7. Related question

Although the gas is preserved in the ice with a slight delay, (i.e., for
the gas "captured" at the surface to reach sufficient depth as to be
trapped in solid ice for later analysis), what correction is made for
the discontinuity caused by a warm period?

e.g., Having seen the pictures of meltwater pools in Greenland & elsewhere,
the water we see there is actually old ice that had been preserved then
subsequently melted. If a cold snap comes in and freezes the pools again
before allowing the usual progression to continue, there will be a marked
unconformity at the boundary. Doesn't this screw with the models?
It would definitely affect the expected timescales of events and would
also affect the maxima as the higher CO2 levels would have been lost
through the melting (due to increased temperature) of the ice containing
them.

:shrug:

Not sure if I'm explaining myself properly here but I'm picturing the
ice strata in the same way that I would rock strata and when there is
a subsequent erosive episode, the normal progression (e.g., of fining
sequences) is interrupted and a new sequence begins at the point when
deposition restarted. If, instead of examining the minerals, grain size
and so on, you were plotting the CO2 content of a series of samples at
stages along the core, how would you cope with the missing data?
How would you even know that there *was* missing data?

[OKIsItJustMe]

8. Well... let's say today's melting will interfere with future ice cores... (if there are any...)

However... today is an unusual situation (to the best of our knowledge.)

Just to follow your line of reasoning though...

Let's see, X years ago, it was warm enough to melt Y years of deposited ice (removing those Y years of warming from the record) Z years of warming and then cooling later it's cold enough to restart deposition.

I'd say there would be a very abrupt change in the record at that point. (Wouldn't you?) A curve of rising temperatures and GHG concentrations followed sharply by a curve of lowering temperatures and GHG concentrations.

(Could that possibly escape notice?)

[Nihil]

9. Part of my concern is that we can't prove how "unusual" it is ... due to the lack of data ...

> Let's see, X years ago, it was warm enough to melt Y years of deposited
> ice (removing those Y years of warming from the record) Z years of warming
> and then cooling later it's cold enough to restart deposition.

To take your example, there is no data from the gap between year (X-Y)
and year (X+Z). No trapped CO2, no methane, no oxygen isotopes or whatever
else you happen to be measuring. The gap of Y+Z years has literally
evaporated from history as presented by that core. In addition, because
both Y and Z are unknowns, we cannot accurately predict either where
year X was nor what temperature/etc. values were present at that time.
That lack of information will affect any models derived from the sequence.

There will be a "step" in levels that depends on the hysteresis but
(as the temperature is now in the "freezing" range again) it will appear
to indicate a maximum point reached rather than a hole in the record.

> I'd say there would be a very abrupt change in the record at that point.
> (Wouldn't you?)

I'm not sure.

If, at time X+Z, there was a dramatic (=cataclysmic) event that dropped
the temperature suddenly from "damn hot" to "freezing again" then there
would certainly be an abrupt change in the record but if the curve on
either side of the true maximum (i.e., in the missing section) was gradual
(as most natural processes over geological time appear to be) then the
point on the curve matching "refreeze" will not be far from the point
on the curve matching "thaw", only the rate of change will be different.

> A curve of rising temperatures and GHG concentrations followed sharply
> by a curve of lowering temperatures and GHG concentrations.

A tipping point? Yes, that much would be visible (though the dating of
it would be suspect and the critical value would be inaccurate).

> (Could that possibly escape notice?)

Depending on the size of the "step" compared to the longer term trends
then it might escape notice (variation within error bars) or it might
be put down as "the turning point" when the actual turning point was
Y years later along the rising curve. This would suggest that the reversing
mechanism is already under way at year (X-Y) whereas it didn't actually
come into play until sometime between (X) and (X+Z).


(Please bear with me: I am *not* turning into some kind of GW denier
but I *am* having difficulty relating my geological knowledge of
unconformities to the corresponding events in ice-cores. Thanks!)

[OKIsItJustMe]

10. Okay, take it back to your rock strata...

What is it about rock strata that caused geologists to think about plate tectonics; or that caused biologists to think about mass extinction events?


In one word, I would say it was "discontinuities."


The melting you're considering would (I feel) create an obvious discontinuity in the data. The only way to determine how long the period between (X-Y) and (X+Z) was would be extrapolation, and it would probably be uncertain extrapolation at best, but I think the discontinuity would be clear, leading to the conclusion that, "About (X-Z) years ago, something dramatic happened..."

[Nihil]

13. Fair enough ...

Thanks for your patience!
:hi:

[kristopher]

11. The areas from which the samples are taken rarely experience snowmelt

The warmest recorded temp in Vostack was -19C. Greenland is less reliable.

http://en.wikipedia.org/wiki/Vostok,_Antarctica
Ice core drilling
420,000 years of ice core data from Vostok, Antarctica research station. Current period is at left.From bottom to top: Solar variation at 65°N due to Milankovitch cycles (connected to 18O). 18O isotope of oxygen. Levels of methane (CH4). Relative temperature. Levels of carbon dioxide (CO2).
420,000 years of ice core data from Vostok, Antarctica research station. Current period is at left.From bottom to top: Solar variation at 65°N due to Milankovitch cycles (connected to 18O). 18O isotope of oxygen. Levels of methane (CH4). Relative temperature. Levels of carbon dioxide (CO2).

In the 1970s The Soviet Union drilled a set of cores 500–952 m deep. These have been used to study the oxygen isotope composition of the ice, which showed that ice of the last glacial period was present below about 400 m depth, Then three more holes were drilled: in 1984, Hole 3G reached a final depth of 2202 m; in 1990, Hole 4G reached a final depth of 2546 m; and in 1993 Hole 5G reached a depth of 2755 m; after a brief closure drilling continued during the winter of 1995. In 1996 it was stopped at depth 3623 m, by the request of the Scientific Committee on Antarctic Research that expressed worries about possible contamination of Lake Vostok. This ice core, drilled collaboratively with the French, produced a record of past environmental conditions stretching back 420,000 years and covering four previous glacial periods. For a long time it was the only core to cover several glacial cycles; but it has recently (2004) been exceeded by the EPICA core, which whilst shallower, covers a longer time span.<2> In 2003 drilling was permitted to continue, but was halted at the estimated distance to the lake of only 130 m.

Although the Vostok core reached a depth of 3623 m the usable climatic information does not extend down this far. The very bottom of the core is ice refrozen from the waters of Lake Vostok and contains no climate information. The usual data sources give proxy information down to a depth of 3310 m or 414,000 years.<3> Below this there is evidence of ice deformation. It has been suggested that the Vostok record may be extended down to 3345 m or 436,000 years, to include more of the interesting MIS11 period, by inverting a section of the record.<4> This then produces a record in agreement with the newer longer EPICA record, although it provides no new information.

[Nihil]

14. OK

That makes sense then!
Thanks.
:hi:

[Indy Lurker]

6. Thanks for the replies and graphs

The only ice cores I was aware of were from EPICA and Vostok.

[hatrack]

12. You also might want to check out Lonnie Thompson's work at Ohio State

His concentration has been tropical ice cores, and his work goes back into the 1970s, I believe. Central Asia, China, S. America, Africa, and so forth.