[gradsusr] Accumulated Precip issues

Jeff Duda jeffduda319 at gmail.com
Thu Dec 10 01:19:29 EST 2020


When your data have a variable time increment, there really is nothing you
can do to account for this in a single control file. Your options are to
either:

1) Write three control files, one for each time increment
2) Be judicious and mathematical when making plots from the control file
using the finest time increment

In my own work, I've gone with the latter. Good luck.

Jeff

On Wed, Dec 9, 2020 at 10:55 PM Mike Manning <weathermon83 at gmail.com> wrote:

> Thanks Jeff,
>
>
>
> That’ll work perfect  - do you know what the best way to be of templating
> these data files in the descriptor file which have 3 time series of hourly,
> 3 hourly and 6 hourly? I was thinking of using the chusb however that still
> requires the files to be in an incremental series. Of course if I use
> apf00_%f3 it’ll stop once it hits hour 54 because then it goes to 3 hourly
> after that.
>
>
>
> Hourly from 1-54, 3 hourly from 57-144, 6 hourly from 150-168
>
>
>
> apf00_001
>
> apf00_002
>
> apf00_003
>
> apf00_004
>
> apf00_005
>
> apf00_006
>
> apf00_007
>
> apf00_008
>
> apf00_009
>
> apf00_010
>
> apf00_011
>
> apf00_012
>
> apf00_013
>
> apf00_014
>
> apf00_015
>
> apf00_016
>
> apf00_017
>
> apf00_018
>
> apf00_019
>
> apf00_020
>
> apf00_021
>
> apf00_022
>
> apf00_023
>
> apf00_024
>
> apf00_025
>
> apf00_026
>
> apf00_027
>
> apf00_028
>
> apf00_029
>
> apf00_030
>
> apf00_031
>
> apf00_032
>
> apf00_033
>
> apf00_034
>
> apf00_035
>
> apf00_036
>
> apf00_037
>
> apf00_038
>
> apf00_039
>
> apf00_040
>
> apf00_041
>
> apf00_042
>
> apf00_043
>
> apf00_044
>
> apf00_045
>
> apf00_046
>
> apf00_047
>
> apf00_048
>
> apf00_049
>
> apf00_050
>
> apf00_051
>
> apf00_052
>
> apf00_053
>
> apf00_054
>
> apf00_057
>
> apf00_060
>
> apf00_063
>
> apf00_066
>
> apf00_069
>
> apf00_072
>
> apf00_075
>
> apf00_078
>
> apf00_081
>
> apf00_084
>
> apf00_087
>
> apf00_090
>
> apf00_093
>
> apf00_096
>
> apf00_099
>
> apf00_102
>
> apf00_105
>
> apf00_108
>
> apf00_111
>
> apf00_114
>
> apf00_117
>
> apf00_120
>
> apf00_123
>
> apf00_126
>
> apf00_129
>
> apf00_132
>
> apf00_135
>
> apf00_138
>
> apf00_141
>
> apf00_144
>
> apf00_150
>
> apf00_156
>
> apf00_162
>
> apf00_168
>
>
>
> *From:* gradsusr [mailto:gradsusr-bounces at gradsusr.org] *On Behalf Of *Jeff
> Duda
> *Sent:* Thursday, 10 December 2020 1:47 PM
> *To:* GrADS Users Forum
> *Subject:* Re: [gradsusr] Accumulated Precip issues
>
>
>
> Mike,
>
> You can use a t-index skip value in the sum function. So you would do
> something like *sum(apcpsfc,t=1,t=72,2*) to sum precip grids from *only
> every other time index* in the control file. This, of course, assumes the
> interval is regular and that the 6-hourly data contains all of the 3-hourly
> data. I think for GFS output this is indeed true, but you should check a
> few time steps just to be sure. Plotting the difference between a few pairs
> of time-consecutive precip fields should help you confirm.
>
>
>
> Jeff Duda
>
>
>
> On Wed, Dec 9, 2020 at 8:30 PM Mike Manning <weathermon83 at gmail.com>
> wrote:
>
> Hi all,
>
>
>
> There are several models which provide GRIB2 data with say 3 hourly until
> x amount of hours and then 6 hourly until the end of that run.  Some models
> even have a tertiary time step where they run hourly, 3 hourly and then 6
> hourly data..
>
>
>
> At the moment for templating, I’m creating a control file for each series
> however this doesn’t seem ideal for accumulated precip.  EG I can use
> sum(apcpsfc, t=1, t=72) for a model that uses hourly from 1-72 hours, but
> what would I do for a model that then goes 3 hourly from 72-144 hours for
> accumulated precip? How would I get the entire 1-144 hours of accumulated
> precip totals? Is there a better way of templating the data so it’s all
> combined or does it need to be kept separate because I’m dealing with
> different time series? I can’t seem to find any way to tally together the
> accumulated precip data once the time series changes from hourly to 3
> hourly (and again for models that have 3 time series such as UKMO)
>
>
>
> Cheers, Mike
>
>
>
>
>
> _______________________________________________
> 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
> _______________________________________________
> 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/20201209/7aa1e12f/attachment-0001.html>


More information about the gradsusr mailing list