Hi Andrew,
It sounds like what you want IS the regridding to a seasonal
climatology time axis. What this operation does is take your 100 years
of data, and at each x,y location forms the average of all the data
from J-F-M of every year, for the first time step of the climatological
data (if you use the standard seasonal_reg axis). All the
April-May-June data from all years is averaged to form the second
average and so on. The result of the operation would have this shape
name
Title I J K
L
RAIN_seas
Precipitation 1:178
1:139 -
1:4
Now, you want different months to be included in your seasons. Just to
learn how it all works, you might try first just computing the averages
from the SEASONAL_REG axis in the climatological_axes data set and look
at the results:
yes? use my_data
yes? use climatological_axes; cancel data climatological_axes
yes? let rain_seas = rain[gt=seasonal_reg@MOD]
yes? shade/L=1 rain_seas ! this is the J-F-M average
yes? shade/L=2 rain_seas ! this is the A-M-J average
You'd want to set the plot options so that the same color levels are
used for all the plots, using the /LEVELS qualifier on your plot
command.
To change the timing of the seasons you need to define your own
seasonal climatological axis. We define climatological years in the
year 0000 - 0001, just as a signal to Ferret to not label the year when
it draws graphics. The axis may lie anywhere in the years 0000-0001 as
long as it is less than or equal to 1 year long. This might be
something like:
yes? DEFINE AXIS/T/UNITS=months/T0=1-jan-0000/EDGES/MODULO
seas_ax = {11, 14, 17, 20, 23}
With the time origin at the start of January, mid-January is 0.5
months, mid-February is 1.5 months and so on. So the start of December
is 11, the boundary between Feb and March is 14, and so on, to the end
of November at month 23.
yes? list t[gt=seas_ax]
VARIABLE : T
axis SEAS_AX
SUBSET : 4 points (TIME)
15-JAN / 1: 12.50000
16-APR / 2: 15.50000
17-JUL / 3: 18.50000
16-OCT / 4: 21.50000
The first point, L=1 will be a (D,J,F) average, then (M,A,M),
(J,J,A), and finally (S,O,N) for L=4. Now, if you regrid your data to
this axis, it'll contain the seasons you want to define.
yes? let rain_seas = rain[gt=seas_ax@MOD]
Ansley
Andrew Cottrill wrote:
Hi guys
I am trying
to calculate the seasonal climatologies
of gridded rainfall data that covers about
100 years
and is in monthly format..
The original
data is like:
name
Title I J K
L
RAIN
Precipitation 1:178
1:139 -
1:1344
I want to
plot maps of average seasonal rainfall for summer(D,J,F)
autumn (M,A,M), winter (J,J,A), and spring (S,O,N).
I have
searched the archives but most work is for
producing time series which I currently don’t want.
The grid
starts from 16-jan-1890:16-dec-2001 and has
1344 time points
I presume
the set climatological_axes etc is for time
series analysis only, as is the modulo gridding
(seasonal_reg,
month_reg and month_irreg).
I am fairly
new at ferret and would appreciate any
suggestions.
Regards
Andy Cottrill
USQ Australia
--
~> ~> ~> ~> ~> ~> ~> ~> ~> ~> ~>
Ansley Manke, NOAA/PMEL ansley.b.manke@noaa.gov
7600 Sand Point Way NE, Seattle WA
Phone 206-526-6246, FAX 206-526-6744
|