Re: time_counter as axis

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

Hi Arthur,
It sounds like you've about got it. I don't see anything wrong
with your commands. Ferret does not interpolate your data to
days as it's computing the monthly climatology; days are the
coordinate units but the monthly data is kept as monthly data.

This example works correctly, just using all of your commands
but just integers as the data.  See if any of the list commands I
have here help you figure out what's happening with your data:

  yes? let v_3to5yr = tsequence(t[t=25:60])
  yes? list/L=1:5  v_3to5yr
             VARIABLE : TSEQUENCE(T[T=25:60])
             SUBSET   : 5 points (T)
   1   / 1:  25.00
   2   / 2:  26.00
   3   / 3:  27.00
   4   / 4:  28.00
   5   / 5:  29.00

  yes? def axis/t=0:1095.7275/npoints=36/edges/units=days/\
     t0="01-jan-1974:00:00" t3to5yr
  yes? let v_3to5yr_t = v_3to5yr[gt=t3to5yr@asn]

  yes? def axis/t=0:365.2425/edges/npoints=12/
     \t0=1-jan-0001/units=days/modulo tclim
  yes? let v_clim_12mon = v_3to5yr_t[gt=tclim@mod]

  yes? list v_clim_12mon
             VARIABLE : V_3TO5YR_T[GT=TCLIM@MOD]
             SUBSET   : 12 points (TIME)
   16-JAN      /  1:  37.00
   15-FEB      /  2:  38.00
   18-MAR      /  3:  39.00
   17-APR      /  4:  40.00
   17-MAY      /  5:  41.00
   17-JUN      /  6:  42.00
   17-JUL      /  7:  43.00
   17-AUG      /  8:  44.00
   16-SEP      /  9:  45.00
   17-OCT      / 10:  46.00
   16-NOV      / 11:  47.00
   17-DEC      / 12:  48.00

     ! Check some averages
  yes? list v_3to5yr_t[l=1:36:12]  ! January values averaged
             VARIABLE : V_3TO5YR[GT=T3TO5YR@ASN]
                        regrid: 365 day on T
             SUBSET   : 3 points (TIME)
   JAN-1974 / 1:  25.00
   JAN-1975 / 2:  37.00
   JAN-1976 / 3:  49.00

  yes? list (28+40+52)/3    ! mean agrees with Jan of v_clim_12mon
             VARIABLE : (28+40+52)/3
          40.00


  yes? list v_3to5yr_t[l=4:36:12]  ! April values averaged
             VARIABLE : V_3TO5YR[GT=T3TO5YR@ASN]
                        regrid: 365 day on T
             SUBSET   : 3 points (TIME)
   APR-1974 / 1:  28.00
   APR-1975 / 2:  40.00
   APR-1976 / 3:  52.00
  yes? list (25+37+49)/3    ! mean agrees with Apr of v_clim_12mon
             VARIABLE : (25+37+49)/3
          37.00



Arthur M. Greene wrote:

Hi Jason,

Thanks for the reply! Here's what I've been able to do (There are 5
years of monthly variable v; we wish to construct a climatology for the
last three years):

let v_3to5yr = tsequence(v[L=25:60])   ! Extract desired time slice

define axis/T=0:1095.7275/npoints=36/edges/units=days/\
T0="01-jan-1974:00:00" t3to5yr

let v_3to5yr_t = v_3to5yr[GT=t3to5yr@asn]

So far, so good. I now have three years of data on a real-time axis,
beginning with jan-1974. The plot looks OK. Now for the climatology:

define axis/T=0:365.2425/EDGES/NPOINTS=12/\
T0=1-JAN-0001/UNITS=DAYS/MODULO tclim

let v_clim_12mon = v_3to5yr_t[GT=tclim@MOD]

This does produce a 12-month plot that _looks_ like a climatology (i.e.,
x-axis reads Jan, Feb, Mar...), but the values are not equal to the
averages of the corresponding month's values in v_3to5yr_t, i.e.,
v(Jan,v_clim_12mon) <> (Sum of Jan vals in v_3to5mon)/3. So perhaps I
don't understand how the MOD (or ASN, for that matter) climatologies are
actually computed. Even if the 36 months of data were interpolated into
daily values, and the appropriate daily values averaged to produce the
monthly climatology, the resultant would not appear to be correct. The
docs do not seem to clarify exactly how these climatologies are computed
(or maybe I need a refresher!)

Your comments greatly appreciated!

Thanks,

Arthur

On Wed, 2004-09-01 at 08:33, Jaison Kurian wrote:

  

! --------------example.jnl--------------------------------
   ! extract last three years data ie. abstract index L=25:60

   let v_3to5yr = TSEQUENCE(v[L=25:60])

   ! define monthly time axis from JAN-1997 to DEC-1999 & assign it
   !     T=0:`365.2425*3`/NPOINTS=36/UNITS=DAYS together will make it for 
   !      3 years and  T0 will take care of actual year stamp.

      define axis/T=0:`365.2425*3`/NPOINTS=36/EDGES/units=days/\
                  T0="01-JAN-1997:00:00" t3to5yr

      let v_3to5yr_t = v_3to5yr[gt=t3to5yr@asn] 

   ! Define a monthly climatological axis 
   
      define axis/T=0:365.2425/EDGES/NPOINTS=12/T0=1-JAN-0001/\
                     UNITS=DAYS/MODULO tclim

   ! Find monthly climatology by "MODULO" regridding

      let v_clim_12mon = v_3to5yr_t[GT=tclim@MOD]
      plot v_clim_12mon     

! If you want to find anomaly or want to plot the climatology
!   over the data then please refer :
!   http://ferret.pmel.noaa.gov/Ferret/Mail_Archives/fu_2003/msg01470.html

!---------------------- end of example.jnl-------------