The Schmidt hammer as a relative-age dating tool and its potential for calibrated-age dating in Holocene glaciated environments

Abstract

The Schmidt hammer is a relatively cheap, portable, sturdy instrument with proven value over the last two decades or so in rapidly dating coarse inorganic deposits of diverse origins. Early views were that its dating role was limited to distinguishing recently exposed from much older. Typically, either a few sites of possibly different ages or occasional older surfaces amongst many young sites were studied. More recently, calibration curves based on individual R-value means from small numbers (2–4) of sites of known ages have been used to estimate the ages of undated sites. We present Schmidt hammer rebound ( R–) values from 28 ‘Little Ice Age’ (and younger), 23 Preboreal and 7 Younger Dryas glaciated surfaces in southern Norway in order, first, to test rigorously the robustness of the instrument as a relative-age dating tool. Despite being obtained from different surfaces (moraines, glaciofluvial deposits and bedrock) and varied metamorphic lithologies, the R-value overall means and 95% confidence intervals for the ‘Little Ice Age’, Preboreal and Younger Dryas age categories (respectively, 60.0±1.6, 41.6±1.4 and 34.2±2.0) are statistically significantly different. Only two outlying sites in the two younger age categories have overlapping confidence intervals, demonstrating remarkable robustness in differentiating early- and late-Holocene surfaces. The distinction between Preboreal and Younger Dryas sites (with terminal dates <2000 years apart) is less clear but still statistically significant, though possibly partly because of enhanced weathering conditions at the predominantly well vegetated Younger Dryas sites. Second, we examine the feasibility and desirability of controlling non-age-related factors, including some previously considered critical (instrument wear, operator bias, initial rock surface texture), which emerge either as less important than previously argued or as relatively unimportant, together with others previously unreported (e.g. long-term changes in lichen, soil, snow and vegetation covers). Third, we investigate the potential for calibrated-age dating by applying exploratory, linear rates of R-value decline to selected combinations of sites. The results suggest that error limits of ca ±700 to ±1600 years should be achievable over the Holocene timescale. This improved dating capability, however, will require adequate numbers of site means not only for each age category used to define these curves but also for each set of test surfaces of the same ages. Recommendations are made for a suitable sampling protocol for developing further the Schmidt hammer as a calibrated-age dating tool.