Project

General

Profile

cdo eca_tn10p (Abort): Input streams have different number of time steps!‏

Added by Anonymous about 14 years ago

Hi All,

I have the t2min-1990s.nc file, is daily min temperature over 1990-01-01 to 1999-12-31 (3652 days -->3652 time step).
I have calculated the 10th percentile values over all time steps of t2min-1990s.nc by used:

#cdo timpctl,10 t2min-1990s.nc -timmin t2min-1990s.nc -timmax t2min-1990s.nc pctl10-t2min-1990s.nc

Then I got pctl10-t2min-1990s.nc, is 10th percentile contained the last time step of t2min-1990s.nc

and then I used eca_tn10p operator:

#cdo eca_tn10p t2min-1990s.nc pctl10-t2min-1990s.nc cold-nights.nc

To calcuclate cold nights percent wrt 10th percentile of reference period.
But I got this error mesege:

"cdo eca_tn10p (Abort): Input streams have different number of time steps!"

What is wrong?

All avantages are appreciate,
-Jo


Replies (2)

RE: cdo eca_tn10p (Abort): Input streams have different number of time steps!‏ - Added by Ralf Mueller about 14 years ago

Sorry for the late response, but I'm not very familiar with these operators. I tried to get in contact with a meteorologist, but her whole group has changed the office building. Bad timing, but I will start some experiments with eca_tn10p and keep looking for her.

regards
ralf

RE: cdo eca_tn10p (Abort): Input streams have different number of time steps!‏ - Added by Ralf Mueller almost 14 years ago

This is a work around for a special case of yearpctl, but I mighyt be possible to to make generalizations. Lets say, you are interested in the number of days with a temperature above 90% of the maximum value.

  1. compute the percentile with
    cdo yearpctl,90 input.nc -yearmin input.nc -yearmax input.nc outputPctl90.nc
  2. For multiple years, split first with input and the percentile
     cdo splityear input.nc inSplit_
    and
     cdo splityear outputPctl90.nc pcrtlSplit_
  3. For each year, perform a substraction of the original data and the percentile
    cdo sub inSplit_2001.nc pcrtlSplit_2001.nc subInPctl90.nc
    and use the operator for the number of consecutive days with temperature above a given value eca_csu:
    cdo eca_csu,0 -addc,273.15 subInPctl90.nc consecDays_2001.nc
    Eca_csu expects Kelvin temperatures, so in case of Celsius 273.15 has to be added. The last 2 cdo calls can be combined into
    cdo eca_csu,0 -addc,273.15 -sub inSplit_2001.nc pcrtlSplit_2001.nc consecDays_2001.nc
    (1-2/2)