6-month northern summer and winter

Hallo,
I have 4 years daily datasets extending from 1990 to 1993. And I just wonder
is there any trick by using CDO to group my data to 6 month northern summer: March to October
and 6 month northern winter November to April. The final result that i need is two group data
which are:
1. Daily data climatology of March - October 1990-1993.--> Northern Summer
2. Daily data climatology of November - April 1990-1993. --> Northern Winter

Thanks any insight.
Thomas

Replies (3)

RE: 6-month northern summer and winter - Added by Jaison-Thomas Ambadan over 11 years ago

Hi,

If your input data sets are separate files; then first merge those data using "mergetime" operator, then use "ydaymean" for the climatology and then"selmon" to select the required months; like:

1. cdo -mergetime my_data_199[0-3].grb all_data.grb
2. cdo -ydaymean all_data.grb all_daily_clim.grb
3. cdo -selmon,3/10 all_daily_clim.grb mar_to_oct.grb
4. cdo -selmon,11,12,1,2,3,4 all_daily_clim.grb nov_to_apr.grb

off-course you can combine the operations 1-3 in one step.

Also for your inf: From an earlier post by Uwe Schulzweida : "it is a good praxis to name/rename your files in a way that there are alphanumerically sorted. it is very important for the CDO operators copy, cat and mergetime."

Hope this helps!

Cheers,
J.

RE: 6-month northern summer and winter - Added by T. Tob over 11 years ago

Hi J,

Thanks, but i got a problem with the time-axis when i executed the command:

[root@localhost MJO_Clivar]# cdo ydaymean V850.nc V850_clim.nc
cdo ydaymean: Processed 19194912 values from 1 variable over 1826 timesteps ( 0.69s )

then I checked the output:

NOTE: Coordinates out of order or missing on axis time at subscript 60*

NOTE: A dummy axis of subscripts will be used

thus, when i tried to simply plot the output i got a abnormal time axis increment 1,2,3,4 that should be
1-Jan-1990, 2-Jan-1990, etc (for daily climatology).

dimensions:
lon = 144
lat = 73
level = 1
time = 366 // unlimited
variables:
float lon ( lon )
standard_name : longitude
long_name : Longitude
units : degrees_east
axis : X

float lat ( lat )
         standard_name :    latitude
         long_name :    Latitude
         units :    degrees_north
         axis :    Y

float level ( level )
         standard_name :    air_pressure
         long_name :    Level
         units :    millibar
         positive :    down
         axis :    Z

double time ( time )
         standard_name :    time
         units :    hours since 1800-01-01 00:00:00
         calendar :    standard

short vwnd ( time, level, lat, lon )
         standard_name :    northward_wind
         long_name :    Daily V-wind on Pressure Levels
         units :    m/s
         add_offset :    187.65
         scale_factor :    0.01
         _FillValue :    -32767
         unpacked_valid_range :    ( -140, 175 )
         actual_range :    ( -98.86, 96.32 )
         precision :    2
         least_significant_digit :    1
         GRIB_id :    34
         GRIB_name :    VGRD
         var_desc :    v-wind
         dataset :    NCEP/DOE AMIP-II Reanalysis (Reanalysis-2) Daily Averages
         level_desc :    Pressure Levels
         statistic :    Mean
         parent_stat :    Individual Obs
         cell_methods :    time: mean (of 4 6-hourly values in one day)
=

Thanks a lot.
Thomas

RE: 6-month northern summer and winter - Added by Jaison-Thomas Ambadan over 11 years ago

Hi,

You could force the time axis using "settaxis" operator, which will solve the problem I guess:

e.g., cdo -settunits,days -settaxis,1990-01-01,00:00:00,1day V850_clim.nc ofile.nc

cheers,
J

(1-3/3)

Project

General

Profile

CDO