Re: [ferret_users] time-axis definition problem

[Ansley Manke <ansley.b.manke@xxxxxxxx>]

Hi Maria,
When you put the :24 in the time specifier

  ,T="01-jan-1980":"30-dec-2000":24]

were you wanting to regrid the data to a new axis?  When you put a third argument there, the :24, that is defining a new axis and making a regridding specification.  See "dynamic axes" in the documentation,
http://ferret.pmel.noaa.gov/Ferret/documentation/users-guide/Grids-Regions/GRIDS#_VPINDEXENTRY_649

I don't know off the top of my head what the details for doing that for a time axis would be.  I think that the problem you're seeing is that the dynamic axis defined by T="01-jan-1980":"30-dec-2000":24] is a Gregorian time axis and therefore can't correctly be combined with the 360-day axis of your data.

If I understand correctly what you are trying to do, I think the best method to put the data from all the monthly datasets onto a single plot would be to explicitly define a long time axis; something like this:

define axis/t-1-jan-1980:30-dec-2000:24/units=hours/calendar=360_day alltime

plot/z=4/vlim=11:21/color=black/line/nolab temp_grid[d=3,x=-11:@ave,y=41:48@ave,gt=alltime]
plot/over/z=4/vlim=11:21/color=black/line temp_grid[d=4,x=-11:@ave,y=41:48@ave,gt=alltime]
...

-Ansley

On 4/3/2012 2:33 AM, Marina Chifflet wrote:

Hi ferreters,

 

I have model outputs on a 360_day calendar (Roms output of scenarios with IPSL forcings).

I have one file per month, for 20 years.

 

I would like to plot a variable (ex SST) from 01-Jan-1980 to 30-Dec-2000 for example (1 month after 1 month).

To do this, I define a time-axis at the same time of my first plot:

 

> plot/nolab/z=4/vlimits=11:21/color=black/line temp_grid[d=3,x=-11:0@ave,y=41:48.5@ave,T="01-JAN-1980":"30-DEC-2000":24]

 

the first variable temp_grid is defined from 01-Jan-1980 to 30-Jan-1980

 

but Ferret answer me an error of time (see attached file):

 

 

indicating that the temp_grid variable is defined from October 2008 to February 2030.

 

If I check with the command

> show grid temp_grid

 

the time definition of temp_grid in January 1980 is confirmed.

And if I plot this variable without defining explicitly the time-axis in the command:

> plot/nolab/z=4/vlimits=11:21/color=black/line temp_grid[d=3,x=-11:0@ave,y=41:48.5@ave]

 

it is plotted in January 1980….

 

 

So it seems that an error is induced by the definition of the time-axis in the “plot” command.

Is the 360_day calendar a problem?

The time axis and grid are previously defined as:

> define axis/calendar=360_day/T="01-JAN-1980:12:00":"30-JAN-1980:12:00":24 taxis_01

> define grid/T=taxis_01/X=xaxis/Y=yaxis/Z=zdepth gridBiscay01

 

And then the variable temp_grid:

> let/d=3/bad=-999 temp_z = IF mask_rho[d=1] LT 1 THEN (-999) ELSE ZAXREPLACE(temp[d=3],depth,z[gz=zdepth])

> let/d=3 temp_grid = temp_z[d=3,g=gridBiscay01@ASN]

 

Do you have any idea to correct this error, or to do it in another way?

 

Thank you very much!

Marina

_____________________________________________

Dr Marina Chifflet

AZTI - Tecnalia / Marine research Division
Herrera Kaia Portualde z/g
20110 Pasaia (Gipuzkoa) - Spain
Tel: 943 00 48 00 (ext. 890) - Fax: 946 572 555
e-mail: mchifflet@xxxxxxxxxxx
www.azti.es ; www.tecnalia.info

www.azti.es | www.alimentatec.com | www.itsasnet.com T. (+34) 94 657 40 00 | F. (+34) 94 657 25 55