#26 closed defect (fixed)
Problem of phasing-out between Tair and Swdown for the WATCH forcing.
Reported by: | nvuilsce | Owned by: | jan.polcher |
---|---|---|---|
Priority: | major | Milestone: | ORCHIDEE 2.0 |
Component: | Driver files | Version: | orchidee_1_9_6 |
Keywords: | Watch; Phasing-out | Cc: |
Description
When running ORCHIDEE with a WATCH forcing file (WFD_1989.nc), the outputs at high frequency (30 minutes) show that there is a problem of phasing-out between Tair and Swdown. See figure below.
As one can see, the maximum of SWdown occurs normally at noon but the maximum for Tair occurs at 6pm and the minimum Tair at 9am. I think that Tair is shifted of one forcing time step (3h) too late.
Remind that the value read at a certain time step t in the Netcdf file corresponds to the instantaneous value at t+1. Consequently, for a forcing file starting the first January at 00:00 (like the WATCH forcing), the first value stored for Tair should correspond to the temperature at 3am.
I think the time-serie for Tair should start one time step late (= the first value should be removed). It is presumably the same for the other instantaneous fields (Wind Speed, Air Pressure and Specific Humidity).
Attachments (2)
Change History (6)
Changed 11 years ago by nvuilsce
Changed 11 years ago by jpolcher
comment:1 in reply to: ↑ description Changed 11 years ago by jpolcher
Replying to nvuilsce:
When running ORCHIDEE with a WATCH forcing file (WFD_1989.nc), the outputs at high frequency (30 minutes) show that there is a problem of phasing-out between Tair and Swdown. See figure below.
As one can see, the maximum of SWdown occurs normally at noon but the maximum for Tair occurs at 6pm and the minimum Tair at 9am. I think that Tair is shifted of one forcing time step (3h) too late.
Remind that the value read at a certain time step t in the Netcdf file corresponds to the instantaneous value at t+1. Consequently, for a forcing file starting the first January at 00:00 (like the WATCH forcing), the first value stored for Tair should correspond to the temperature at 3am.
I think the time-serie for Tair should start one time step late (= the first value should be removed). It is presumably the same for the other instantaneous fields (Wind Speed, Air Pressure and Specific Humidity).
Jan
As shown by the analysis of the forcing data of WFD and ERA-Int it is clear that it is the SWDOWN of WFD which is out of phase. As the time axis of the original data received from the Hadley Centre were not changes, it is very likely that this problem has its origin in the processing which produced the WFD data during the WATCH project.
Graham Weedon will be contacted to see if this is a known problem.
comment:2 Changed 7 years ago by nvuilsce
This has been futher discussed in the wiki page: http://forge.ipsl.jussieu.fr/orchidee/wiki/Branches/Driver_Improvements
This is now accounted for internally in the new driver.
comment:3 Changed 7 years ago by nvuilsce
- Resolution set to fixed
- Status changed from new to closed
comment:4 Changed 7 years ago by jpolcher
Discussed and solved here : https://forge.ipsl.jussieu.fr/orchidee/wiki/Branches/Driver_Improvements
Tair and Swdown half-houlry values 'seen' by ORCHIDEE with a WATCH forcing file.