The Great Greenland Snow Job – 08 – The Delta 18O Dating Debacle

The Delta 18O Dating Debacle

One of the enduring mysteries surrounding the GRIP ice core is how [roughly] 3 metres of annual precipitation at Summit Camp manages to become just 23 centimetres of annual core ice when the monthly average high surface temperature always remains well below freezing.

Summit Camp Climate

GRIP Depth Age Scale from Flow Modeling

It is based on a Dansgaard-Johnsen (1969) type flow model, with a 1500 m lower shear layer, 13.5% bottom sliding and accumulation rates which are a function of the d18O values: 23.0 cm of ice at -35.2 per mil with 8% and 18% change per per mil at -35.2 per mil and -40 per mil, respectively.

Investigating this mystery leads directly to the counting of annual ice layers that have been identified by establishing the ice core d18O values.

The d18O analysis of the GRIP 93 pit [diagram below] clearly shows a relationship between temperature and the d18O values [for this particular site].

The minimum d18O values usually spike downwards in the December-January winter period reaching between -40 and -45.

The maximum d18O values usually spike upwards in later summer reaching between -29 and -25.

The d18O values are variable [like the weather] and the plotted values display: shoulders, double peaks, double troughs and periods of very rapid change between the extremes.

However, it should be noted that:

1) This is a rounded view that is not based upon daily data.
2) The clarity of the d18O signal is based upon [about] 50 data points per year.
3) The clarity of the temperature signal is based upon only [about] 23 data points per year.

GRIP 93 Pit

Greenland and North Atlantic climatic conditions during the Holocene – as seen in high resolution stable isotope data from Greenland ice cores.
Bo Møllesøe Vinther – Ph.D. Dissertation – 2006
University of Copenhagen

Obtaining similar high resolution d18O results from the GRIP ice core, for example, would require a sampling increment of 4.6 millimetres [23 cm per year / 50] which roughly translates to 217 samples per metre.

Sadly, the resolution of the d18O GRIP ice core data that is available for download has an unbelievably low resolution of 550 millimetres which equates to 1 sample every 2.39 years [based upon the headline accumulation rate of 23 centimetres per year].

GRIP Oxygen Isotopes

This data set contains 55cm average delta 18-O values from the GRIP core.

Clearly, the ultra-low resolution d18O GRIP download data can mask a multitude of sins.

Similarly, the low resolution of the GRIP 93 pit temperature plot masks a very variable weather pattern that is revealed by the GISP2 Automatic Weather Station.

GISP2 Weather

GISP2 Automatic Weather Station Daily Averages
Stearns, C R; Weidner, G A (2000)

Although the basic summer and winter peaks are evident in the GISP2 daily temperature data there is also a lot of noise that blurs the simplistic low resolution patterns presented by the GRIP 93 pit graphic.

The volatile nature of the intervening noise indicates that temperature spikes during the wintertime could easily be confused for summertime peaks in the d18O record.

Therefore, the more reliable method of counting years from the d18O record is based upon counting wintertime peaks – it’s just like counting tree rings [and just as subjective as dendrochronology].

This point is underlined in the following example from the Crete ice core where the associated text clearly states the winter layers define the annual layers.

Crete d18O sample
Ten meters of stable isotope data from the Central Greenland ice core Crête containing 19 annual layers. The dashed lines indicate the winter layers and define the annual layers.

University of Copenhagen – Centre for Ice and Climate

Armed with this understanding of how d18O tree rings are counted lets take a look at how the chronology was established for the 1973 Milcent ice core which sings harmoniously from the mainstream 1259 volcanic event hymn sheet.

Greenland Hymn Sheet

An Inter-Hemisphere Volcanic Time-Marker in Ice Cores from Greenland and Antarctica
C.C. Langway, Jr., H.B. Clausen and C.V. Hammer
Annals of Glaciology 10 1988 – International Glaciological Society

The Milcent d18O data presents a perfect pristine snow job chronology [without depth data] which provides “8 samples per year” based upon “an annual accumulation of 55.6 cm/year ice” i.e. a low resolution of just over 14 samples per metre.

d18O Maximum.: -21.51
d18O Minimum..: -37.64
d18O Average….: -29.60697851
d18O Mid-Point..: -29.575
d18O Range……: 16.13

Milcent Snow Job


Zooming in on the period 1760 to 1800 [for example] in the Milcent ice core provides an insight into how the chronology reflects the d18O tree rings.

Using very low thresholds values of Average Minus One to identify winter values and Average Plus One to identify summer values we find:
a) Years that are all summer: 1770, 1798
b) Pairs of years with no summer: 1787+1788 and 1783+1784
c) Summer peaks in December and January: 1762/3/4/5/6/7/8/9, 1774, 1780/1/5/6 and 1795/6/7/8/9
d) Years that are all winter: 1792
e) Winter peaks at midyear: 1762/6, 1778/9, 1781 and 1797
f) Years with no winter: 1760/5/8/9 and 1798

Milcent 1760-1800

Clearly, the chronology and the d18O tree rings are not aligned and the Milcent chronology contains many phantom [phoney] years between 1761 and 1800.

Further insight into this period can be gained from the manual records which have a resolution of about 28 samples per metre i.e. twice the resolution provided in the download.

Milcent station d18O - manual 1761-1880
Stable isotopes and climate history from polar ice cores – Thorsteinn Thorsteinsson
NASA Astrobiology Institute – University of Hawaii

Remarkably, the manual records don’t establish thresholds values around the average to distinguish between winter and summer values – they just use the average value as their benchmark.

Any d18O value rising above the average is deemed to represent a summer and even getting close to the average is enough to qualify as a summer in some instances.

Similarly, d18O values dipping just below the average are deemed to represent a winter.

Furthermore, in this surreal world of d18O tree rings they appear to be using the less reliable summer values as their primary annual marker [highlighted in black in the original].

Thus, a visual inspection of the manual record reveals at least 40% of the years in the chronology between 1761 and 1800 are probably phantom years and that the height of the annual ice layers appear to increase [on average] below [about] 70 metres.

Unfortunately, the manipulation of the d18O data doesn’t end there because when the data is deemed too noisy they have ways of making it talk softly.

The investigation of d18O records from the upper part of the ice cores is complicated by water vapor diffusion in the pores of snow and firn in the top 60-70 m of the ice sheet.

This diffusion will tend to smooth the high frequency oscillations of the d18O.

In Figure A.2 and A.3 the diffusion related decay of the annual d18O cycle can be seen for a central Greenland shallow core.

Diffusion Corrected Data

Measurements of d18O from different depths in the Site E central Greenland ice core.
Left: near the surface before firn diffusion dampens the annual cycle.
Right: below the diffusion zone, the annual cycle is significantly dampened.
The thin curve is diffusion corrected data.

Using the Johnsen et al. (2000) diffusion model and the mathematical formulation of the diffusion problem presented in Johnsen (1977), it is possible to correct d18O series for diffusion.

Having diffusion corrected data, the underlying seasonally resolved climatic signal can be retrieved from ice core d18O records.

Greenland and North Atlantic climatic conditions during the Holocene – as seen in high resolution stable isotope data from Greenland ice cores.
Bo Møllesøe Vinther – Ph.D. Dissertation – 2006
University of Copenhagen

And when the d18O data is whispering they have ways of making it talk loudly.

Back Diffusion Restores Data

Stable isotopes and climate history from polar ice cores – Thorsteinn Thorsteinsson
NASA Astrobiology Institute – University of Hawaii

Therefore, if you prefer your data to talk without a painful accent, you are best advised to reset your watch to 1977 and revisit the truly extraordinary Camp Century 1977 Annual Layer Thickness and Delta 18-O chronology where the annual ice layers begin to thicken below 70 metres and the data suggests the ice sheet formed around 945 AD.

Camp Century d18O Chronology 1977 - Annual Layer

Camp Century d18O Chronology 1977 - Annual Layer Below 70m

Camp Century d18O Chronology 1977

Camp Century 1977 Annual Layer Thickness and Delta 18-O

Gallery | This entry was posted in Catastrophism, Greenland, History, Inventions and Deceptions. Bookmark the permalink.

2 Responses to The Great Greenland Snow Job – 08 – The Delta 18O Dating Debacle

  1. If one careens the earth to a different axis of spin to, say, having Greenland at an equatorial latitude, and thus changing its climate, then oxygen isotope theory ) becomes a wee bit problematical.

    Has the earth careened during the past? Apparently Herodotus, quoting Egyptian priests, mentioned that the sun often rose where it used to set. Scientifically that would be a primary observation. Unscientifically it would be regarded as a myth or error in dogma.

    These days being scientific is regarded as a heresy.

  2. Pingback: Heinsohn and The Eclipse Record | MalagaBay

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s