[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

20020215: strange problem with precip-type composite



Robert,

presumably you have redefined LLMXGD to 5250000 in MCHPRM.xxx and gemprm.h.

What have you set LLMDGG to in GEMPRM.PRM?

The calculation heap has to be big enough to hold the input fields necessary
from the calculation.

Steve Chiswell
Unidata User Support

On Fri, 15 Feb 2002, Robert Mullenax wrote:

> Let me clarify..the 1800x1200, 1800x700.etc are the kxky params
> in gdradr.
>
> Thanks,
>
> Robert
>
> -----Original Message-----
> From: Robert Mullenax [mailto:address@hidden]
> Sent: Friday, February 15, 2002 8:42 AM
> To: 'address@hidden'
> Subject: strange problem with precip-type composite
>
>
> I have been successfully making U.S. and regional precip-type radar
> composite
> maps for some time using a modified form of Steve Chiswell's script that
> uses
> the RUC grids to estimate sfc and low level temps to determine precip type.
>
> I modified GEMPAK to use up to 5,250,000 grid points (as big as Sun Fortran
> 77
> will let me without breaking gdbiint) and this has worked great for the last
> 6 months.
>
> However, when I started using my regional hi-res grids (south-central U.S.,
> 1800x1700)
> to produce precip-type instaed of just contouring echoes I get snow showing
> up where it should
> not (South Texas).  It is fine on the lower res U.S. composite (1800x1200
> for the whole U.S.).
>
> The thing that seems to make the difference is the RUC file that I use..the
> current RUC file
> will produce the fals snow echoes even though the RUC shows temps way too
> warm for that,
> while I can pick another RUC time and it works fine.   When I use gdlist to
> look at the
> snow parameter when this happens there are a few numbers and then the rest
> is just
> ******  like an overflow has occurred.  However there is no error issued in
> the script
> and also this doesn't happen every time.
>
> I know this is a convoluted question, but does anyone have any ideas?
>
> Thanks,
> Robert
>
> Robert Mullenax
> Weather Systems Administrator
> Universal Weather and Aviation
>