(back to Science Work Robin)
Gravity Wave Drag
(pre-Jan’11) Gravity wave drag ought to affect the vertical temperature profile - Manoj pointed out that the tunable params in there are set for HadCM3 and may be too weak for the lower resolution FAMOUS. Looking at it, the “low level” boundary is almost certainly too high in std. FAMOUS. However, playing the gravity wave drag param doesn’t do very much. Ramping it up pretty high slightly improves the overestimation of the northern hemisphere midlatitude jet, maybe. GWD parameters have thus been left as for HadCM3, rather than spend time and effort tuning something that appears to have only a minor effect.
(newer)It seems, in fact that Manoj didn’t go far enough initially - it seems that, at FAMOUS resolution, the GWD doesn’t really function at all and the model develops unphysical winds in the top level that have some nasty side-effects for the whole climate. He suggested using uniform Rayleigh friction instead, which improves things hugely (even turning the GWD off) although doesn’t give the hemispheric constrast to the jets that a correct GWD parameterisation would - the precise values for the friction and the GWD are the subject of current tuning.
The GWD param requires measures of the subgridscale slopes (dh/dx)^2, (dh/dy)^2 and (dh/dx)*(dh/dy) . It turns out that at UM4.5 these are assumed to have been derived from a 10 minute resolution dataset. For my paleo runs, I don’t have ice-sheet data at this fine a resolution - ICE5G came on 1 degree grids, and Phillipe’s stuff is on 50km polar grids, interpolated to a 1 degree regular grid. Test runs of modern orography ancils derived from ETOPO2 at 10min or 1deg resolution show few differences, the largest being clearly co-located where the use of a lower resolution original set has resulted in a lower mean orography in steep areas. All differences between 10min and 1deg runs were smaller than the differences between the ETOPO2 10min run and whatever was used to derive the original FAMOUS orography ancil, again where the mean orography differed. In conclusion, I don’t think the use of 1degree paleotopographies to derive FAMOUS ancils will be a significant source of error.