Hi Erik,
WRFHelp is correct that this is an NCL issue. The '_10' designation actually refers to the GRIB "time range indicator" (see http://www.nco.ncep.noaa.gov/pmb/docs/on388/table5.html). It actually indicates only that forecast times greater than 255 require 2 bytes to be fully represented. The convention of adding the indicator as a suffix was originally implemented for time range indicators that were otherwise unsupported by NCL. The usage of indicator 10 is now common enough that it should become fully supported. As a workaround for now you will need to read your files into two file variables and concatenate the variables using a common name after you read them. The TimePeriodSuffix option relates to a different issue. I will create a ticket for this problem and hopefully it will be fixed soon. -dave
On Sep 26, 2011, at 11:48 AM, Erik Noble wrote:
> Dear NCL,
> I don't know exactly where to begin with this.
> I am having trouble using the setfileoption & TimePeriodSuffix option to read grib data from the WRF UPP post-proccessor. I initially contacted WRFhelp about this; after several emails, WRFhelp suggested I contact NCL to resolve this. ( The conversation is below. The script I am using is below too.)
>
> Basically, I am using ncl's "addfiles" function to read in several variables from 96 grib files. Everything works just fine, up to the 86th file, where the name of the variable, regardless of the variable I ask for, has a "_10" appended on the end of the variable name. For example, I need to plot U and V winds at isobaric levels. When I try to extract the U variable, which reads as U_GRD_GDS1_ISBL from ncl_filedump, I notice that after
> forecast hour 258 (from wrfout file 2006-09-12-18) it "changes to a new name, "U_GRD_GDS1_ISBL_10."
>
> The WRFhelp team claims this is an NCL issue but at the same time suggested I try using the setfileoption & TimePeriodSuffix. I still get the same result, the last 10 files will not read.
>
> How do I resolve this. Further details towards my converstaion with wrfHelp and my code is below.
> Thank you in advance for your help.
> -Erik Noble
>
>
> On Sep 26, 2011, at 10:29 AM, WRFHelp wrote:
>
> Hey Erik-
>
> I have to admit I have never used setfileoption. But I would first
> check -
> Are value 86+ still in the variable with the _10 appended to the end?
> If so
> you can fire a question to ncl_help and ask why. You may need to read both
> values and swap at the byte roll over time if NCL doesn't support the
> suppression
> of renaming the variable. If you think the data in the
> unipost grib output file is wrong please send me at least one input
> and output
> file (preferably fhr 258 or beyond). You are correct to send me (FTP)
> the file you will
> have to rename it removing the ":", sorry.
>
> If you get this working with NCL I would be interested in how you
> achieved it. Thanks
> and best of luck on the journey.
>
> --wrfhelp
>
> On Fri, 23 Sep 2011 18:26:54 -0500 "Noble, Erik U. (GSFC)[COLUMBIA UNIVERSITY]" <erik.u.noble@nasa.gov>
> wrote:
>
> Hi. Ok. I see. I followed your suggestion.
> I used the set file option
> setfileoption("grb","TimePeriodSuffix",False) before reading in the
> files, as the link suggests.
>
> But, the scripts stopped at the 85th file. There are 96 files (or
> rather 96 values that are supposed to be there).
> Is there something I am doing wrong?
>
> NCL Script
>
> ;***************************************************************
> ; Load Libaries
> ;***************************************************************
> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_code.ncl"
> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_csm.ncl"
> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/contributed.ncl"
> load "$NCARG_ROOT/lib/ncarg/nclscripts/wrf/WRF_contributed.ncl"
>
> begin
> ;***************************************************************
> ; Change these parts
> ;***************************************************************
> DATADir = "./" ;input directory
> Files = "WRFPRS_*" ;input files
> ncdiro = "./" ;output directory
> NCFILE1 = "pressure" ;output file
> ;***************************************************************
> ; Read in WRF Files
> ;***************************************************************
> FILES = systemfunc (" ls " + Files) ; file paths
> numFILES = dimsizes(FILES)
> print(numFILES)
>
> setfileoption("grb","TimePeriodSuffix",False)
>
> files = addfiles (FILES+".grb", "r")
> ;========================
> ListSetType (files, "join") ; concatenate "cat" or merge
> "join"
> ; z = files[:]->HGT_GDS1_ISBL ;Geopotential height
> ; t = files[:]->TMP_GDS1_ISBL
> ; td = files[:]->DPT_GDS1_ISBL
> u = files[:]->U_GRD_GDS1_ISBL
> ; v = files[:]->V_GRD_GDS1_ISBL
> ; sph = files[:]->SPF_H_GDS1_ISBL
> ; rh = files[:]->R_H_GDS1_ISBL
>
> printVarSummary(u)
> printMinMax(u,True)
> print(u(:,0,0,0))
> end
>
> Terminal Output
> [eunoble@discover18] post_Experiment_01 $ ncl test_pressure.ncl
> Copyright (C) 1995-2011 - All Rights Reserved
> University Corporation for Atmospheric Research
> NCAR Command Language Version 6.0.0
> The use of this software is governed by a License Agreement.
> See http://www.ncl.ucar.edu/ for more details.
>
>
> Variable: numFILES
> Type: integer
> Total Size: 4 bytes
> 1 values
> Number of Dimensions: 1
> Dimensions and sizes: [1]
> Coordinates:
> (0) 97
>
> Variable: u
> Type: float
> Total Size: 1051117800 bytes
> 262779450 values
> Number of Dimensions: 4
> Dimensions and sizes: [ncl_join | 86] x [lv_ISBL3 | 25] x [g1_lat_0
> | 311] x [g1_lon_1 | 393]
> Coordinates:
> lv_ISBL3: [50..1000]
> g1_lat_0: [-18.402..35.335]
> g1_lon_1: [324.251..35.749]
> Number Of Attributes: 12
> center : US National Weather Service - NCEP (WMC)
> long_name : u-component of wind
> units : m/s
> _FillValue : 1e+20
> level_indicator : 100
> gds_grid_type : 1
> parameter_table_version : 2
> parameter_number : 33
> model : North Pacific Hurricane Wave Model
> forecast_time : 0
> forecast_time_units : hours
> initial_time : 09/02/2006 (00:00)
> (0)
> (0) u-component of wind: min=-58.0981 max=48.6898
>
>
> Variable: u (subsection)
> Type: float
> Total Size: 344 bytes
> 86 values
> Number of Dimensions: 1
> Dimensions and sizes: [ncl_join | 86]
> Coordinates:
> Number Of Attributes: 15
> g1_lon_1 : 324.251
> g1_lat_0 : -18.402
> lv_ISBL3 : 50
> initial_time : 09/02/2006 (00:00)
> forecast_time_units : hours
> forecast_time : 0
> model : North Pacific Hurricane Wave Model
> parameter_number : 33
> parameter_table_version : 2
> gds_grid_type : 1
> level_indicator : 100
> _FillValue : 1e+20
> units : m/s
> long_name : u-component of wind
> center : US National Weather Service - NCEP (WMC)
> (0) -0.2010425
> (1) 0.5233667
> (2) 1.252354
> (3) 2.798071
> (4) 4.349082
> (5) 2.420659
> (6) 0.4866016
> (7) 1.312463
> (8) 2.137537
> (9) 3.360134
> (10) 4.575947
> (11) 5.127134
> (12) 5.67635
> (13) 5.544646
> (14) 5.411767
> (15) 5.634192
> (16) 5.861868
> (17) 6.084302
> (18) 6.315991
> (19) 5.298777
> (20) 4.281072
> (21) 3.456113
> (22) 2.63082
> (23) 2.895886
> (24) 3.170952
> (25) 4.251602
> (26) 5.347261
> (27) 4.80581
> (28) 4.269766
> (29) 4.413843
> (30) 4.56592
> (31) 4.179683
> (32) 3.794978
> (33) 4.186904
> (34) 4.583442
> (35) 4.875342
> (36) 5.170044
> (37) 4.927932
> (38) 4.673596
> (39) 4.54365
> (40) 4.401719
> (41) 5.855395
> (42) 7.312293
> (43) 8.062078
> (44) 8.802468
> (45) 7.902993
> (46) 7.010733
> (47) 6.265562
> (48) 5.535437
> (49) 6.774873
> (50) 8.025661
> (51) 8.683907
> (52) 9.33523
> (53) 8.577312
> (54) 7.812954
> (55) 7.792734
> (56) 7.772768
> (57) 7.932007
> (58) 8.090359
> (59) 8.378562
> (60) 8.665559
> (61) 6.949731
> (62) 5.23176
> (63) 7.035208
> (64) 8.830354
> (65) 8.120037
> (66) 7.405535
> (67) 7.409487
> (68) 7.408035
> (69) 6.654251
> (70) 5.898555
> (71) 6.32137
> (72) 6.728426
> (73) 7.347017
> (74) 7.962986
> (75) 6.482971
> (76) 5.008667
> (77) 4.458303
> (78) 3.911145
> (79) 4.192146
> (80) 4.46812
> (81) 4.676645
> (82) 4.890271
> (83) 5.270593
> (84) 5.647607
> (85) 5.126626
>
> On Sep 23, 2011, at 10:25 AM, WRF Help wrote:
>
> Excellent --
>
> OK -- why? Well the GRIB format originally only had 8 bits for
> representing the forecast hours. So you ran out of numbers at 255 (decimal). Well today in the GRIB PDS we set byte 21 (1 relative) to 10 to indicate we need to use bytes 19 AND 20 to represent P1 (the time). So in your UPP GRIB output at forecast hour 258 the time representation in the GRIB file changes. NCL looks at these
> values and when it sees the time representation change it changes the
> variable name. See http://www.ncl.ucar.edu/Document/Manuals/Ref_Manual/NclFormatSupport.shtml#GRIB
> for more details about that. Unfortunately in your case this is not a new variable just a format dependency.
>
> Now I have never tried this but it sounds like it might work. In NCL use the setfileoption function to turn off the TimePeriodSuffix (not available in NCL until 5.1.0).
>
> Hope that helps -- otherwise assume both names are the same variable.
>
> --wrfhelp
>
>
> On Wed, 21 Sep 2011 17:57:24 -0500 "Noble, Erik U. (GSFC)[COLUMBIA UNIVERSITY]" <erik.u.noble@nasa.gov<mailto:erik.u.noble@nasa.gov>> wrote:
>
> Dear WRF HELP,
> I have 96 WRF output files (every 3 hours, 288 forecast hours total.)
> I ran unipost in your UPP. It created the files WRFPRS_d01.000 to WRFPRS_d01.288. So far so good.
>
> But when I actually try look at the new data itself, using NCL's ncl_filedump command, the last 10 WRFPRS
> files are missing data.
> The only way I can prove this is to show you the results from the ncl_filedump command script that prints out the value for every single time-step. The last 10 files are missing, but the actual WRF output files are full of data. What happened?
>
>
> The variables change.
>
> I need to plot U and V winds at isobaric levels. When I try to
> extract the variable I need, U_GRD_GDS1_ISBL, I notice that after
> forecast hour 258 (from wrfout file 2006-09-12-18) it "changes to a
> new name, "U_GRD_GDS1_ISBL_10."
> It is not that the last 10 files are missing data, it is that the
> variable names have changed. Why would this occur?
> See below.
> -Erik
>
> NCL Script
>
> ;***************************************************************
> ; Load Libaries
> ;***************************************************************
> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_code.ncl"
> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_csm.ncl"
> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/contributed.ncl"
> load "$NCARG_ROOT/lib/ncarg/nclscripts/wrf/WRF_contributed.ncl"
>
> begin
> ;***************************************************************
> ; Change these parts
> ;***************************************************************
> DATADir = "./" ;input directory
> Files = "WRFPRS_*" ;input files
> ncdiro = "./" ;output directory
> NCFILE1 = "pressure" ;output file
> ;***************************************************************
> ; Read in WRF Files
> ;***************************************************************
> FILES = systemfunc (" ls " + Files) ; file paths
> numFILES = dimsizes(FILES)
> print(numFILES)
>
> setfileoption("grb","TimePeriodSuffix",False)
>
> files = addfiles (FILES+".grb", "r")
> ;========================
> ListSetType (files, "join") ; concatenate "cat" or merge
> "join"
> ; z = files[:]->HGT_GDS1_ISBL ;Geopotential height
> ; t = files[:]->TMP_GDS1_ISBL
> ; td = files[:]->DPT_GDS1_ISBL
> u = files[:]->U_GRD_GDS1_ISBL
> ; v = files[:]->V_GRD_GDS1_ISBL
> ; sph = files[:]->SPF_H_GDS1_ISBL
> ; rh = files[:]->R_H_GDS1_ISBL
>
> printVarSummary(u)
> printMinMax(u,True)
> print(u(:,0,0,0))
> end
> _______________________________________________
> ncl-talk mailing list
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Mon Sep 26 12:21:05 2011
This archive was generated by hypermail 2.1.8 : Mon Sep 26 2011 - 14:45:50 MDT