[gradsusr] Generating Ensemble Probabilities Based on Conditional Threshold(s)

[Jeff Duda]

Aaron,
I presume the error occurs during this command:

'define probp01 =
100.0*ave(const(const(maskout(precip,precip-0.01),1),0,-u),e=1,e=30)'

Is that correct?

If so, there could be a discrepancy in the structure of the ensemble
dimension in the GEFS files. I'm pretty sure that G(E)FS precipitation is
stored as an alternating series of 3- and 6-hour precip accumulation
intervals. So during forecast hours that are divisible only by 3 and not 6
(i.e., 3, 9, 15, 21 ...) the GRIB2 identifier for APCPsfc may be different
than at the other hours.

What I recommend is running g2ctl on two different valid time GEFS files to
see if anything about the output (other than the valid time in the file
changes). If it does, then I bet that's what it is, and you'll have to use
more clever ways of templating over the member and forecast dimensions.

Jeff Duda

On Tue, Jun 1, 2021 at 2:52 PM Perry, Aaron <
axp09171 at alumni.northernvermont.edu> wrote:

> Good Afternoon,
>
> I am trying to use the GFS Ensemble to generate probabilities based on
> conditional thresholds and am running into some trouble.
>
> Initially, I have tried to generate an ensemble mean of the probability of
> 0.01" of QPF in 3-hourly time increments.
>
> See attached GrADS script (gefs_p01.gs) and the control file (gefs.ctl)
> that I am using.
>
> I am receiving the following error messages:
>
> *Operation error:  Invalid dimension environment*
> *  Looping dimension does not vary*
> *result is Operation error:  Invalid dimension environment*
> *  Looping dimension does not vary*
>
> What am I doing wrong here?
>
> Best,
>
> *Aaron Perry*
> Meteorologist
> Class of '16
> Northern Vermont University-Lyndon
> (formerly Lyndon State College)
> Boston, Mass.
> Twitter: @AaronPerryWx <https://twitter.com/AaronPerryWx>
> <https://twitter.com/arnpry> <https://twitter.com/AaronPerryWx>
> Email: Aaron.Perry at lyndonstate.edu
> _______________________________________________
> gradsusr mailing list
> gradsusr at gradsusr.org
> http://gradsusr.org/mailman/listinfo/gradsusr
>


-- 
Jeff Duda, Research Scientist
University of Colorado Boulder
Cooperative Institute for Research in Environmental Sciences
NOAA/OAR/ESRL/Global Systems Laboratory
Boulder, CO
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://gradsusr.org/pipermail/gradsusr/attachments/20210601/47fd8106/attachment-0001.html>