Several field studies yield values of n < 1 at stresses between ca. 0.001 and ca. 0.1MPa (e.g. Meier, 1960; Marshall etal., 2002), consistent with either diffusion creep (n = 1, p = 2-3) or Harper-Dorn creep (n = 1, p = 0). Experimentally constrained estimates of the diffusion creep rate, however, suggest that diffusion creep of glaciers and ice sheets is unlikely for grain sizes > 1 mm and stresses > 0.0001 MPa over the entire range of glaciologically significant temperatures (Goldsby & Kohlstedt, 2001). Moreover, stresses in the diffusion creep regime are too low to activate the dislocation sources required to develop strong crystallographic fabrics, which are nearly ubiquitous in glaciers and ice sheets (Frost & Ashby, 1982). The operation of Harper-Dorn creep (a Newtonian dislocation mechanism that is independent of grain size) in glaciers is inconsistent with field studies, which demonstrate that flow of glaciers and ice sheets is grain size-sensitive (e.g. Cuffey et al., 2000a; see below). Finally, I note that values of n = 1 from field studies are in many cases only apparent values if grain size increases with depth (stress) in glaciers, as is commonly observed. The low apparent value of n results from the decrease in strain rate with increasing grain size as depth (stress) increases, which is typically not taken into account in analyses of mechanical data from field studies.
Was this article helpful?