Rock surface exposure age dating using cosmogenic nuclides

Sample 1 (dolerite) --  showing three of the four faces

Sample 1 -- reverse side, showing the smooth and iron-stained "secondary weathering surface"

This is the latest technique, now pretty well tried and tested, which will have to be used on a substantial scale if we are ever to sort out the length of time over which the Stonehenge bluestones have been exposed to the atmosphere.  There is a vast reference list, but the introductions written by Bethan Davies are very useful.  Here is one of them:

https://www.antarcticglaciers.org/glacial-geology/dating-glacial-sediments-2/cryospheric-geomorphology-dating-glacial-landforms/cosmogenic-nuclide-dating-cryospheric-geomorphology/exposure-age-sampling-methodologies/

Whether one ius sampling a glaciated rock surface or a glacially-transported boulder, one should be aware that even within one small area there may be very different "nuclide accumulation" histories, as illustrated here:

This illustration comes from a paper about avalanches and landslides, but it can equally well be applied to glacial landscapes where block entrainment and prolonged abrasional processes have altered surface profiles.  Erratics or fallen blocks on a moraine or in a glacial trail may have very different TCN concentrations and hence different apparent "cosmogenic exposure ages".  A sloping or vertical surface will also have a different exposure to incoming radiation than a horizontal surface, and will therefore have a younger "exposure age."

Getting reliable dates from samples is no simple matter, and over and again in the literature we see results that have to be rejected simply because the samples taken have included surfaces with very different surface exposure ages and also buried quartz fragments with inherited ages.  In some of the samples I have been taking (with permission!) and looking at, there are 4 or 5 surfaces, each with a different thickness of weathering crust and different degrees of iron staining.  This is because when you try to take a sample from a horizontal surface (as recommended by the experts) the rock fractures in complex ways because of the stress exerted at the impact point of the hammer or chisel. Fractures will occur on pre-existing planes of weakness such as joints and faults. If any of these surfaces have been exposed to cosmic bombardment, the exposure age will be altered.  

Also, because cosmic bombardment can influence a rock surface to a depth of 10 m or more, if abrasion and surface fracturing on an erratic or bedrock exposure during a glacial episode only extends to a depth of 5 cm (for example) then the "new" surface will carry with it an inherited age.  This will lead to an over-estimation of the time that has elapsed since the last glacial episode.

Note also that cold ice protects, and warm ice erodes -- so it is possible for a prolonged glaciation to have very little effect on the subglacial rock surface, with little or no abrasion or plucking and the preservation of the old TCN signal......

Two examples.  

In sample 1 (unspotted dolerite), at the head of this post, the sample obtained from a good bash with the hammer has four faces.  Obviously, if one is trying to obtain a reliable age for the oldest weathered face (with a good crust and some lichen traces) it is not a good idea to incorporate in the processed sample a sizeable part of the secondary weathered and iron-stained face, because the nuclide signature here will be different;  and so the aggregate date will be younger than it should be.  But here comes the problem.  If the rock you are dealing with does not contain much quarz, and you just shave off the primary weathered surface to a depth of 10mm and throw the rest away, you may not get a large enough sample for the lab to deliver a reliable result.  For some rock types they may need a sample weighing 1 kg or more -- so you know in advance that your delivered date will be an aggregate date, and not very reliable.

In sample 2 (spotted dolerite), there are 5 surfaces, two of which are fresh and unweathered and 3 are weathered to varying degrees.  Again, if the laboratory sample is shaved off to a depth of maybe 10mm to incorporate just the rough primary weathered surface, a reasonably accurate result will be obtained.  On the other hand, if the whole sample chunk is needed in order to provide adequate quartz for the lab analysis, then the weathered (and possibly exposed) secondary and tertiary surfaces will have to be incorporated, introducing more errors and compromising the accuracy of the end result.  Even bigger errors will be introduced if the rock beneath the surface carries an inherited age from earler exposure to the atmosphere and cosmic bombardment.

Two images of sample 2 (spotted dolerite)

In general, dolerite is not a very good rock for cosmogenic nuclide dating, because it is not rich in quartz.    Granites and sandstones are apparently much more suitable........

From Google searches:

Inherited ages, or inheritance, in rock surface dating refer to an older-than-actual age recorded by a rock surface due to a previous exposure to cosmic rays that wasn't completely "zeroed" by sufficient erosion or signal resetting. This is a problem in cosmogenic nuclide (CN) dating, where the accumulation of nuclides is interrupted but not fully reset. Inheritance can lead to an overestimation of the rock's last true exposure or burial event. 

Rock surface exposure age inheritance is when a rock contains an older-than-actual surface age, often because cosmogenic nuclides from previous periods of surface exposure were not fully removed by subsequent erosion or burial. This "inherited component" complicates accurate dating, especially in glacial settings where cold-based ice can shield rocks from erosion, leading to apparent ages that are too old. Geologists account for inheritance by sampling multiple boulders, looking for ages that are tightly clustered, and identifying outlier ages as likely having inherited components.