Re: D_TIME_UNIT (monthly axis time encoding)

To: BROCKMANN_Patrick <patrick@xxxxxxxxxxxxxxxxxxxxxxx>
Subject: Re: D_TIME_UNIT (monthly axis time encoding)
From: hankin@xxxxxxxxxxxxxxxxxxx (Steve Hankin)
Date: Thu, 20 Nov 1997 08:42:42 -0800
Cc: ferret_users@xxxxxxxxxxxxxxxxxxx
Encoding: 2 TEXT BOUNDARY, 110 MESSAGE, 2 TEXT BOUNDARY, 112 MESSAGE, 3 TEXT BOUNDARY
In-reply-to: BROCKMANN_Patrick <patrick@jungle.POLYTECHNIQUE.FR> "D_TIME_UNIT in a multi netCDF descriptor file" (Nov 20, 1:03pm)
References: <34742721.41C6@jungle.polytechnique.fr>
Sender: owner-ferret_users

		*** HELP WITH FERRET MONTHLY AXES ***

		see also discussion in the FAQ at
    http://tmap.pmel.noaa.gov/ferret/FAQ/data_management/equal_months.html

--------------------------

Hi Patrick,

A "month" is always a problematic unit, since it is ambiguous. The best
solution is to use a well-defined unit such as an hour (i.e. D_TIME_UNIT=3600
seconds) and then encode your months in terms of the number of hours. I will
attach the descriptor file coads_climatology.des that illustrates this
approach.

If you already have a collection of NetCDF files that have embedded in them a
time step encoding of 1.0, 2.0, 3.0, ... representing month number then the
easiest solution is fix the data set on-the-fly in a Ferret script using

	use climatological_axes			! defines axis  "MONTH_REG"
	cancel data climatological_axes		! axes are persistent

	set data my_MC_data
	LET/D=my_MC_data var1_corrected = var1[GT=MONTH_REG@ASN]
	.
	.
	.

Put these lines into a script called "my_MC_data.jnl" and then use
"yes? GO my_MC_data" instead of "yes? use my_MC_data" -- the variables
var1_corrected (etc.) will have the time encoding that you want.

Note that Ferret does have a unit "month" defined. It is 2628000 seconds (730
hours) ... 1/12 of a 365 day year (for modeling purposes). It is not a good
idea to build a modulo climatological axis from this axis since it will "drift"
significantly off over many years due to the absence of leap years.

	- steve

P.S. The bug you found earlier in the L=12 point of the climatological axis has
been fixed in version 4.9 -- expect to see it early next week!

========================================================================

On Nov 20,  1:03pm, BROCKMANN_Patrick wrote:
> Subject: D_TIME_UNIT in a multi netCDF descriptor file
> Dear ferreters,
>
> Thank you for the trick to get the last timestep of
> a climatological axes (shade/L=12 clim[L=11:12]).
>
> But I still have a problem using this time climatological axes
> and my own Multi netCDF.
>
> The problem comes from the unrecognition of my time axis which
> is not read as monthly.
> To correct this, I'd like to know what is the exact value of
> the namelist entity called D_TIME_UNIT in the descriptor file if
> I want to represent months.
>
> I have already tried 2592000s (for a month of 30 days),
> 2629746s (for a month of 730.485 hours), 2635200 (for a month of 30.5
> days)
> but none of them are recognized when I open my multi netCDF
> and I look to the time axis.
>
> **************************************************
> yes? show grid/T AM_29
>     GRID GDV11
>  name       axis              # pts   start                end
>  LON       LONGITUDE          360mr   179.5W               179.5E(539.5)
>  LAT       LATITUDE           180 r   89.5S                89.5N
>  normal    Z
>  TIME8     TIME                24 r   17-JAN-1988 19:46    09-DEC-1989
> 22:53
>
>            L     T                 BOX_SIZE       TIME_STEP ()
>            1>  17-JAN-1988 19:46:40  1              1056.5
>            2>  16-FEB-1988 22:00:00  1              1057.5
>            3>  18-MAR-1988 00:13:20  1              1058.5
>            4>  17-APR-1988 02:26:40  1              1059.5
> ...
>
> --> Nothing is written in the parenthesis after TIME_STEP
>
> **************************************************
>
> Thanks again
>
> --
> Patrick BROCKMANN - Laboratoire de Meteorologie Dynamique
> Groupe d'Analyse du Rayonnement Atmospherique
> Ecole Polytechnique - Route Departementale 36
> 91128 Palaiseau Cedex - FRANCE
> Tel: 01 69 33 31 91 - Fax: 01 69 33 30 05
> mailto:Patrick@Jungle.Polytechnique.Fr
> http://Jungle.Polytechnique.Fr
>-- End of excerpt from BROCKMANN_Patrick

-- 

		|  NOAA/PMEL               |  ph. (206) 526-6080  
Steve Hankin	|  7600 Sand Point Way NE  |  FAX (206) 526-6744
		|  Seattle, WA 98115-0070  |  hankin@pmel.noaa.gov

1 ********************************************************************************
 *        NOAA/PMEL Tropical Modeling and Analysis Program, Seattle, WA.        *
 ********************************************************************************
 $FORMAT_RECORD
 D_TYPE  = '  GT',
 D_FORMAT        = '  1C',
 D_SOURCE_CLASS  = 'OBSERVED DATA    ',
 D_SOURCE        = ' ',
 D_SUBSOURCE     = 'SURFACE MARINE',
 D_PROG_REV      = '0.00'
 $END
 $BACKGROUND_RECORD
 D_EXPNUM        = 'COAD',
 D_MODNUM        = 'MNCL',
 D_TITLE 	 = 'COADS Monthly Climatology (1946-1989)',
 D_MOD_TITLE     = ' ',
 D_FORCING_SET   = '???             ',
 D_T0TIME        = '1-JAN-0000 00:00:00',
 D_TIME_UNIT     =   3600.000    ,
 D_TIME_MODULO   = .TRUE.,
 D_NTEGRATE_TSTEP =   1.000000    ,
 D_NTUPLE        =           1,
 D_TIME_RUN      = '1-JUL-1991',
 D_WHO_RANIT     = 'Jerry Davison',
 D_TIME_MADE     = '1-JUL-1991',
 D_WHO_MADEIT    = 'Jerry Davison',
 D_AUX_MODNUM    = 6*'    ',
 D_ADD_PARM      = 6*'      ',
 d_grid_filename = 'coads.grd',5*' '
 $END
 $MESSAGE_RECORD
 D_MESSAGE       = ' ',
 D_ALERT_ON_OPEN = F,
 D_ALERT_ON_OUTPUT = F
 $END
 $VARIABLE_RECORD
 D_VAR_CODE      = 'SST',
 D_VAR_TITLE     = 'SEA SURFACE TEMPERATURE',
 D_VAR_TITL_MOD  = '                                        ',
 D_VAR_UNITS     = 'Deg C',
 D_GRID_NAME     = 'COADS',
 D_MISSING_FLAG  = -1.0e34,
 D_BAD_FLAG      = -1.0e34,
 D_ORDERING      = 'WE', 'SN', 'UD', 'TI',
 D_GRID_START    =          1,          1,           1,           1,
 D_GRID_END      =        180,         90,           1,          12,
 D_PRECISION     = 'SINGLE      ',
 D_AUX_SET_NUM   =           0
 $END
 **************************************************
 $VARIABLE_RECORD
 D_VAR_CODE      = 'AIRT',
 D_VAR_TITLE     = 'AIR TEMPERATURE',
 D_VAR_UNITS     = 'DEG C'
 $END
 **************************************************
 $VARIABLE_RECORD
 D_VAR_CODE      = 'SPEH',
 D_VAR_TITLE     = 'SPECIFIC HUMIDITY',
 D_VAR_UNITS     = 'G/KG'
 $END
 **************************************************
 $VARIABLE_RECORD
 D_VAR_CODE      = 'WSPD',
 D_VAR_TITLE     = 'WIND SPEED',
 D_VAR_UNITS     = 'M/S'
 $END
 **************************************************
 $VARIABLE_RECORD
 D_VAR_CODE      = 'UWND',
 D_VAR_TITLE     = 'ZONAL WIND',
 $END
 **************************************************
 $VARIABLE_RECORD
 D_VAR_CODE      = 'VWND',
 D_VAR_TITLE     = 'MERIDIONAL WIND',
 $END
 **************************************************
 $VARIABLE_RECORD
 D_VAR_CODE      = 'SLP',
 D_VAR_TITLE     = 'SEA LEVEL PRESSURE',
 D_VAR_UNITS     = 'MB'
 $END
 **************************************************
 $VARIABLE_RECORD
 D_VAR_CODE = '**END OF VARS**'
 $END
 **************************************************
 $EXTRA_RECORD
 $END
 **************************************************
 $STEPFILE_RECORD
 s_filename      = 'coads_climatology.001' ,
 S_AUX_SET_NUM   =           0,
 S_START =   366.0 ,
 S_END   =   8401.336 ,
 S_DELTA =   730.485 ,
 S_NUM_OF_FILES  =           1,
 S_REGVARFLAG    = '     ',
 S_INDEX = '                       '
 $END
 * UNIX precision bug... S_END   =   8401.335 ,
 **************************************************
 $STEPFILE_RECORD
 s_filename = '**END OF STEPFILES**'
 $END
 **************************************************

References:
- D_TIME_UNIT in a multi netCDF descriptor file
  - From: BROCKMANN_Patrick

Previous by thread: D_TIME_UNIT in a multi netCDF descriptor file
Next by thread: Line types in plot

[Thread Prev][Thread Next][Index]

Dept of Commerce / NOAA / OAR / PMEL / TMAP