GlaciMod

The post-glacial uplift gives us important information about the Earth's rheology, both the lithosphere and the mantle. The viscosity of the mantle can be estimated by calibrating theoretical uplift with the observations. In the calculations below we have varied the viscosity profile while the maximum uplift rate is kept at 8.5 mm/yr (as is observed; cf. figure below).

Observed present rate of uplift (Vestoel, 2006). The red line is the zero-line, conour interval is 1 mm/yr.

A lot of possible mantle viscosity profiles turn out to be unrealistic. This include a uniform viscosity mantle (viscosity 0.65 x 10²¹ Pa s; cf. figure below) and a two-layer mantle with uniform viscosity in the upper (UM) and lower mantle (LM), respectively. A two-layer model with uniform UM and LM is promoted by Peltier, Lambeck and others. An example of a two-layer mantle (upper mantle viscosity of 1.5 x 10²¹ Pa s and lower mantle viscosity of  2 x 10²¹ Pa s) is shown below. It is, however, possible to get a better fit with the observations by increasing the elastic thickness of the lithosphere, but this turns out to give significant mismatch with the sea level observations (e.g. in western Norway) over the last 12 000 years.  

Theoretical present (glaci-isostatic) rate of uplift with a uniform model and a two-layer UM/LM model

The only possibility to get a match with the observations is to introduce a low-viscosity asthenosphere in the uppermost part of the mantle. The best-fitting model is shown in the figure below. Based on observations of the present rate of uplift we can conclude that the Earth's mantle is generally of low viscosity (not much higher than 1x10²¹ Pa s) overlain by a 75-150 km thick asthenosphere with much lower viscosity. 

Theoretical present rate of uplift with a mantle viscosity of 1x10²¹ Pa s
beneath a 75 km low-viscosity asthenosphere of 1.8 x 10¹⁹Pa s.