Actually I have worked out how you could do it.
You would have to do some none dimensionless analysis so that you could normalise the rate of decent against the other variables.
A bit like using reynolds number in the wind tunnel.
Its 15 years since I have done any of that in anger so your on your tod.