Alison has just pointed out that Colin Hines has a new paper on his Doppler Spread Parameterisation. The abstract states that:

An embryo Lagrangian DST is introduced and employed to assess the original DST. Earlier results near the Eulerian spectral peak are found to be reasonably valid, whereas those at greater vertical wavenumber are confirmed to have produced too much spreading. The earlier DSP is found to need little if any change, though specific values are suggested for its two most important "fudge factors".

The two numbers he gives are Φ1 = 1.9 and Φ2 = 0.4. These are suggested in order to get suitable values of the Lagrangian Richardson number, to compare with observational values. As it happens, these are precisely the values which I have used in my previous papers (e.g. height of model lid influences on stratospheric climate, forcing planetary waves in the mesosphere, and simulating the quasi-biennial oscillation). It’s also obviously good news that his parameterisation is standing up to more rigorous Lagrangian analysis.