saving daily and 6-hourly data

[rnicholas@atmos_washington_edu]

Just wanting confirm that I have things set up correctly before I dive in to a new set of runs...

I'm running CAM with the data ocean model at T42 and would like to save the following variables at the following intervals:

• monthly: default variables + total precipitation
• daily: sea level pressure, total precipitation, and temperature at all 26 levels
• 6-hourly: zonal and meridional winds at 200 and 850 mb

Furthermore, suppose I want each of my history files to contain exactly one year of data.

So, if I understand the documentation correctly, I would need to set namelist variables as follows:

nhtfrq = 0,-24,-6
mfilt = 12,365,1460
fincl1 = 'PRECT'
fincl2 = 'PSL', 'PRECT', 'T'
fincl3 = 'U200', 'U850', 'V200', 'V850'

Does this look right to folks? If not, can anyone suggest what I ought to change to get the output I'm looking for?

Thanks,
Rob

 

[rneale]

I think that looks ok. I'm not quite sure about the monthly mean files (fincl1). I have never tried to write more than one time sample to the h0 file. I am not sure whether mfilt(0) is hard wired to 1, I fear it may be.

 

[rnicholas@atmos_washington_edu]

Thanks for the response...

According to section B.13 in the user's guide, all elements of MFILT default to 30, so I'm hoping this will be fine.

I noticed that you mentioned MFILT(0) in your reply but the documentation seems to indicate that the first element of such namelist variables have an index of 1. Is there such a thing as MFILT(0)? My sense, from some tinkering around with short test runs, is that MFILT(1) corresponds to history series 'h0', MFILT(2) to history series 'h1', and so on.

~Rob

 

[rnicholas@atmos_washington_edu]

Well Rich, the user guide notwithstanding, it appears you were right -- I'm still getting monthly h0 files regardless of the value I give MFILT(1). MFILT(2) and MFILT(3) work as expected, of course.

Any chance this can get fixed in the documentation?

 

[rneale]

I think idea was to force that on the h0 files so that the standard diagnostics package could be used and that the output would be every real month and not every 30 days fro example. If you want the output structure you were looking for you could use the nco operators to concatenate all the h0 files into one file for the year.