CDO

I have already corrected it but it seems still unrealistic value for the precipitation. This are the commands:
First I substract my ensemble RCM projection (pr_djf.nc) by the E-OBS data (djf_rr_newdom.nc).
After that I'll divide the result with the E-OBS again and multiply it with 100 for making it in percentage.

cdo sub pr_djf.nc djf_eobs_rr_newdom.nc sub_bias_pr.nc
cdo mulc,100 -div sub_bias_pr.nc djf_eobs_rr_newdom.nc bias_pr_djf.nc

The values are not realistic.If anyone can help please ! Thank you

hm, I am nor sure if both physical quantities are comparable

• djf_eobs_rr_newdom.nc holds an amount of rain in mm
• pr_djf.nc hold a precipitation flux in mm per 450sec

I can see in the history that you multiplied this by the daylength in seconds (-mulc,86400) and reset the unit.

So all-in-all the values are not reliable for me. Further more you show in your table some error numbers which cannot be checked with files you uploaded. your uploads have only a single filestep whereas you show errors for different seasons.

In case your error number are in % these number might be fine. I cannot check the original values myself, so it;s very hard to comment on that

cheers
ralf

Hello Ralf!
Thank you for the response. The pr_djf.nc data is from 5 ensemble RCM simulation. It is daily data that's why I multiply it with 86400.
But my question is it in mm per 450sec, how can I convert it into mm/daily?
Thank you very much

In essence you best check the simulation code or ask the developers if possible. I see two options

1. the unit is correct (flux in mm per 450sec)
   1. in this case data can a be instantaneous values: then your are lost because only 450 sec of the whole day is covered or
   2. it is a mean values: then you might multiply it with 192 (=86400/450) to get to the complete flux of the day
2. the unit is not correct
   1. the values are already accumulated for a day and someone forgot about correcting the unit: in that case you can use the plain values for further daily analysis
   2. the rest is too much guessing - go back and check the model

best wishes
ralf