[ Read Responses | Return to the Index ]

[ Previous | Previous in Thread | Next ]

Message: Re: Initializing model with RUC analysis

Posted by Greg Thompson on 27 Jan 00, 10:15 MT, in response to Initializing model with RUC analysis, posted by Will Cheng on 6 Jan 100, 12:10 MT

The best reason to include real obs into your own meso model
initialization is if you plan to run a different resolution
than the RUC (vertically as well as horizontally). My personal
opinion is the RUC does an excellent job with initial fields by
not smoothing too much (its predecessors smoothed more than I
cared for). This is particularly true in the vertical with
strong inversions - RUC seems to maintain them better than other
model's initialization (check out Bill Moninger's Java applet that
lets you plot RUC soundings vs. RAOBS and look for strong inversions
like in TX/OK today). This improvement is to be expected with 40
levels (all above ground unlike Eta) and so many so near/within
the boundary layer.

If you plan to run the same res as RUC and don't have a decent
raw data quality control package, then you might want to use RUC
fields directly. If, however, you are running a different resolution
you could benefit from doing another analysis given RUC's first
guess gridded data and input RAOBs, METARs, profilers, ACARS, etc.
Back in my CSU-RAMS days, I'd use MAPS for this purpose but then
rely on RAMS to get me a higher quality analysis over my nested grid(s)
with 1/4th the grid spacing of the coarse domain grid.

Good luck.

The FSL RAP/HRRR/RUC Information Forum is maintained with a variant of WebBBS .