Date: Sat, 28 Apr 2001 14:29:50 -0700 Reply-To: houze@atmos.washington.edu Organization: University of Washington X-Accept-Language: en To: Sandra Yuter Subject: Re: qr content Can you provide me with the No, density, and fallspeed vs diameter relations that you are using (with units incicated)? If so, I can take a shot at deriving the beta factor for each. Off hand I don't see why they should be so different. The No, density, and fallspeed vs diameter relations are the only things that are different. Of course the collection efficiencies could differ, but I can't justify anything other than assuming they are both unity. Also, for sanity's sake, can you check and make sure your autoconversion formula is working right. It should be working the same way above and below the 0 deg C level. It should get less potent with height. Also, can you tell me what your vertical velocity profile is? Do you assume it has a constant value at all times? You know, what's really wrong with the class problem formulation is that it applies only at cloud top (no precip falling into the parcel from above).That's why we need to reformulate the problem. Just off the top of my head--you could let this parcel be the cloud top parcel. Then run a second parcel for which the precip falling in from above is that computed for the cloud top parcel. This would give you a fall-in term as well as a fall-out term for both qr and qi. Good luck Date: Mon, 30 Apr 2001 16:11:38 -0700 Reply-To: houze@atmos.washington.edu Organization: University of Washington X-Accept-Language: en To: Sandra Yuter Subject: Re: fall in plots These look interesting. Now, tell me how deep is your column of w=constant? It looks like you are letting the updraft extend all the way up to some great height. Why not let it drop off to zero at some reasonable height? For example, assume a parabolic profile of w with a max at 4 km and going to zero at z=1 and 7 km. This would knock the top of the ice max down to something reasonable with the max ice in the right place. This would be a much more realistic profile than w=constant with height.