This article is taken from:
Davies, B.J., 2022. CRYOSPHERIC GEOMORPHOLOGY: Dating Glacial Landforms I: archival, incremental, relative dating techniques and age-equivalent stratigraphic markers. (link)
Introduction
Schmidt-hammer exposure-age dating assesses the compressive strength of bedrock and boulder surfaces and relates this to the degree of weathering of a surface (Wilson et al., 2019). It is a low-cost relative-dating technique that can be applied rapidly and easily to large numbers of samples in specific locations, allowing for relative ages to be proposed for adjacent features. It has been used in a wide range of Quaternary environments, including glacial landforms (Barr et al., 2017; Shakesby et al., 2006; Winkler, 2014).
Schmidt-hammer dating can complement lichenometric dating and guide more expensive cosmogenic nuclide exposure-age dating sampling strategies for boulders on moraines (Wilson et al., 2019).
Testing rock hardness
Schmidt hammers carry out in situ, non-destructive tests of material hardness (Goudie, 2006). The hammer measures the distance of rebound of a controlled impact on a rock surface. The ‘N’ type Schmidt hammer is most frequently used in geomorphological research. It has compressive strength of 20 to 250 MPa.
When the Schmidt hammer is pressed against a surface, a piston is automatically released onto the plunger. Part of the impact energy is absorbed by the plastic deformation of the rock and part is transformed into heat and sound. The remaining energy represents the “impact penetration resistance” or hardness of a surface (Goudie, 2006). The distance travelled by the piston after rebound is the “rebound value” (R-value).
R-values
The R-value range is 10-100 and is read from a scale on the side of the instrument. Harder rocks have higher R values and are by inference less weathered. R values, therefore, reflect the structural weakening and chemical breakdown of near-surface rock, and increased surface roughness caused by variable resistance to weathering of surface minerals (Shakesby et al., 2006).
This approach can be used to separate moraines formed in different neoglaciations, such as the “Little Ice Age” and older Holocene glaciations (Ffoulkes and Harrison, 2014; Matthews and Owen, 2010; Shakesby et al., 2006; Winkler, 2014) and more recently, even Late Glacial moraines (Barr et al., 2017; Tomkins et al., 2016, 2018a; Wilson et al., 2019) and on blockfields (Marr et al., 2018).
Sampling procedures
- R-values should be obtained from horizontal surfaces free from moss and lichen (Shakesby et al., 2006).
- Samples should avoid the edge of boulders, open and weathered joints, and fissures in the rock.
- Only stable surfaces should be sampled, and sample surfaces should be dry. The microclimate can impart significant variability, and late-lying snow patches, in particular, should be avoided (Goudie, 2006).
- Sampling boulders that have surfaces more than 0.5 m above ground level is recommended (Wilson et al., 2019).
- In any one area, only the same rock types should be sampled, as lithological differences and surface roughness strongly affect R values.
Take many measurements
In general, multiple measurements should be obtained and the mean value recorded. Practitioners should present the mean value and the 95% confidence interval for each site sampled, separated according to surface type or lithology (Shakesby et al., 2006) (Figure 3). Statistical methods such as Students’ T-Test may be used to differentiate between R value populations from different sites.
Shakesby et al. (2006) found little difference in mean R values from particular surfaces in sample sizes of 50, 100, or more readings. This means there may be little advantage in collecting very large samples. Some practitioners discard outliers (Goudie, 2006). Very low values could be due to the rock being weakened by the impact of the hammer, or small rock flaws that were not immediately obvious (Goudie, 2006).
How many measurements?
The number of impact measurements taken per boulder varies among practitioners. Ffoulkes and Harrison (2014) used five samples on 40 boulders per moraine, giving a sample size of 200 readings per moraine. Wilson et al. (2019) sampled a single R value from 250 boulders per site, using the same operator, the same hammer, and in dry conditions. The hammer was periodically tested against the manufacturers test anvil. The mean R value for each site was then calculated.
Matthews and Owen (2010) identified 30 sites at each locality for Schmidt-hammer assessment of bedrock. At each site, 25 R values were recorded from the bedrock surface. The mean R value for the locality was a mean value of the 750 individual measurements.
Barr et al. (2017) measured 30 R values from three surfaces or boulders on each moraine or site, from numerous horizontal positions on each surface. To assess Schmidt hammer drift during sampling, the same granite boulder was sampled at the beginning and end of the sampling period.
Calibration and age conversion
Some operators have attempted to use cosmogenic nuclide ages to calibrate Schmidt hammer R values to obtain numerical ages for glacially transported boulders (Tomkins et al., 2018b, 2016; Winkler, 2009). Calibration curves should normally be considered local, and for a specific lithology (Marr et al., 2018; Shakesby et al., 2006). Multiple tie points are needed across the calibration curve, and a regression can then define the R value-age relationship.
Terrestrial cosmogenic nuclide ages can provide the tie points for linear age-calibration curves and a multiproxy approach to assigning ages to moraines (Wilson et al., 2019; Winkler, 2009). Tomkins et al. (2016) found a statistically significant relationship between 25 granite boulders dated with cosmogenic nuclide dating and Schmidt hammer rebound values, suggesting a linear weathering rate over significant spatial scales for regions of similar climate.
Calculation of numerical ages requires calibration of Schmidt hammer R values, against a calibration boulder or independently dated surface (Dortch et al., 2016). However, this approach has been criticized, as it does not differentiate between instrument calibration and conversion of R values into numerical age information (Winkler and Matthews, 2016).
Online calculators
Tomkins et al. (2018b) provide an online calculator (http://www.shed.earth/shedcalc/) that distinguishes between Schmidt hammer drift following use (instrument calibration) and the development of age calibration curves. They provide an updated age-calibration curve based on 54 independently dating granite boulders (R2 = 0.94, p = < 0.01) (Tomkins et al., 2018b). The calibration curve dataset extends from 0.8 to 23.8 ka.
Quality assurance protocols
- Schmidt Hammer data should be collected by a single operator using a single Schmidt Hammer.
- They should have a clearly defined sampling strategy.
- Samples should only be compared to those of the same lithology and should be taken from flat surfaces more than 0.5 m above ground level.
- Schmidt Hammers should be tested and calibrated against the manufacturers’ anvil before use, after use to test for drift, and ideally periodically during use.
- Multiple measurements (>30) should be taken from each surface and a mean value should be computed.