Time steps in the history files outputs in CAM6

[nuvolet]

Greetings,

I was wondering if somebody could help me to understand the reason of some outputs I got using CAM6 in CESM2.2.2 (running in Derecho).
I asked CAM6 to write frequencies for each history files series each 30 minutes.

FCnudge compset and MERRA-2 reanalysis are used for these tests.
In the namelist (besides other definitions) for the user_nl_cam these are the settings:

avgflag_pertape = 'I'
nhtfrq =  1 # --- hourly output
mfilt  =  48
fincl1 = 'AODVISdn', 'T', 'Q'
model_times_per_day    = 48
nudge_times_per_day    = 8

When I look a the history files T and Q are available each 30 minutes, while AODVISdn are only represented each hour, missing each 30 intermediate tilmestep.

Any comment and/or suggestion will be appreciated.
Thanks in advance

 

[peverley]

Hi Toni,

A couple of things:

1, nhtfrq > 0 is defined as # of timesteps (https://www2.cesm.ucar.edu/models/cesm1.0/cesm/cesm_doc_1_0_4/x2602.html), so if you want hourly output, you should do nhtfrq = -1. That said, this will give you hourly output, not every 30 minutes. You can get (close to) 30 minute output by using whatever number of timesteps is closest to 30 minutes for your run.

2. nhtfrq and mfilt are arrays so you may want to try:

mfilt = 1, 48
nhtfrq = 0, -1

 

[peverley]

Oops - in the example I gave, you'd also want to use fincl2 instead of fincl1

 

[nuvolet]

Thanks Courtney