20010517: cape, bulk richardson

>From: address@hidden
>Organization: UCAR/Unidata
>Keywords: 200105172100.f4HL08p11126

>
>Hi Steve,
>        I'm noticing that too many of the CAPE values are coming up as zero.
>Actually, there were "division by zero" flags raised during the calculations,
>but since a lot of the spring/summer values seemed correct (over 1000 j/kg), I
>  
>didn't think it was a problem. 
>        I've been checking individual measurements with the sounding plots  
>(and index values) for the same date/time/station available on the web.  Sever
> al
>are not agreeing.  For exp, for one of my output times, 5 of the indices showe
> d
>strong instability (LI= -8 , etc.) while the CAPE and Bulk Richardson # 
>were zero.  So I checked it with the sounding plot, the 5 other indices 
>were right on, but the CAPE should have been 3000 j/kg for that day/time.
>
>Is this just resulting fr. missing data in certain soundings?  Any idea
>what's wrong or what can be done?
>
>Thanks,
>Diana
>


Diana,

Indicies like CAPE are sensitive to delta Z values (when integrating 
vertically), so it is possible that a dz <=0 would occur in a sounding
that had duplicate levels. Other indicies that are merely differences at 
several levels are not as likely to encounter those problems.

Other problems can occur, if surface values were missing, dewpoint missing,
or the sounding is truncated prematurely so that the upper bound cannot be 
calculated.

Note that CAPE uses a surface layer of 500 meters by default when
doing calculations (adjustable with the !xxx feature). Using
the layer for lifting the cape parcel is more realistic than just
lifting the surface parcel. 3000 would be huge.

If you send me the SNLIST output of a particular sounding in question,
I'd be happy to look at it tomorrow.

Steve Chiswell
Unidata User Support