Closed StephenGriffies closed 7 years ago
Isn't there a one-line xml entry to produce the monthly from daily?
-- Dr Alistair Adcroft (Alistair.Adcroft@noaa.gov) Princeton University Tel: (609) 987-5073 NOAA/GFDL, 201 Forrestal Road, Princeton, NJ 08540
On Mon, Oct 2, 2017 at 10:20 AM, Stephen Griffies notifications@github.com wrote:
We are saving daily output for the straits data Sections for CMIP6/OMIP
"ocean_Barents_opening", 1, "days", 1, "days", "time" "ocean_Bering_Strait", 1, "days", 1, "days", "time" "ocean_Davis_Strait", 1, "days", 1, "days", "time" "ocean_Windward_Passage", 1, "days", 1, "days", "time" "ocean_Denmark_Strait", 1, "days", 1, "days", "time" "ocean_Drake_Passage", 1, "days", 1, "days", "time" "ocean_English_Channel", 1, "days", 1, "days", "time" "ocean_Faroe_Scotland", 1, "days", 1, "days", "time" "ocean_Florida_Bahamas", 1, "days", 1, "days", "time" "ocean_Fram_Strait", 1, "days", 1, "days", "time" "ocean_Iceland_Faroe_V", 1, "days", 1, "days", "time" "ocean_Iceland_Faroe_U", 1, "days", 1, "days", "time" "ocean_Indonesian_Throughflow", 1, "days", 1, "days", "time" "ocean_Mozambique_Channel", 1, "days", 1, "days", "time" "ocean_Pacific_undercurrent", 1, "days", 1, "days", "time" "ocean_Taiwan_Luzon", 1, "days", 1, "days", "time" "ocean_Agulhas_section", 1, "days", 1, "days", "time" "ocean_Gibraltar_Strait", 1, "days", 1, "days", "time" "ocean_Iceland_Norway", 1, "days", 1, "days", "time"
CMIP6 wants months. Hence, we should change the above first instance of "days" to "months". Sections for CMIP6/OMIP
"ocean_Barents_opening", 1, "months", 1, "days", "time" "ocean_Bering_Strait", 1, "months", 1, "days", "time" "ocean_Davis_Strait", 1, "months", 1, "days", "time" "ocean_Windward_Passage", 1, "months", 1, "days", "time" "ocean_Denmark_Strait", 1, "months", 1, "days", "time" "ocean_Drake_Passage", 1, "months", 1, "days", "time" "ocean_English_Channel", 1, "months", 1, "days", "time" "ocean_Faroe_Scotland", 1, "months", 1, "days", "time" "ocean_Florida_Bahamas", 1, "months", 1, "days", "time" "ocean_Fram_Strait", 1, "months", 1, "days", "time" "ocean_Iceland_Faroe_V", 1, "months", 1, "days", "time" "ocean_Iceland_Faroe_U", 1, "months", 1, "days", "time" "ocean_Indonesian_Throughflow", 1, "months", 1, "days", "time" "ocean_Mozambique_Channel", 1, "months", 1, "days", "time" "ocean_Pacific_undercurrent", 1, "months", 1, "days", "time" "ocean_Taiwan_Luzon", 1, "months", 1, "days", "time" "ocean_Agulhas_section", 1, "months", 1, "days", "time" "ocean_Gibraltar_Strait", 1, "months", 1, "days", "time" "ocean_Iceland_Norway", 1, "months", 1, "days", "time"
Otherwise, we need to write postprocessing code to properly handle the monthly averages, which seems unnecessary and will take more debugging time than we have now. I see no reason to keep daily output when CMIP6 wants months.
Andrew Shao is writing the refineDiag script assuming we will have monthly output. Please make the change to the diag table accordingly.
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/NOAA-GFDL/MOM6-examples/issues/176, or mute the thread https://github.com/notifications/unsubscribe-auth/AFlo8073Eq_lEXlML-cWqG76KufX8My_ks5soPFKgaJpZM4Pqtk- .
For other diagnostics, we are relying on the diag_table.
Why do we need to have dailies for the straits switched to monthlies in the xml rather than making the change in the diag_table?
Yes, we have current analysis based on daily output, even though not requested in CMIP6. Rather than add both daily and monthly entries in the diag_table, it is simplest to use one line in the XML as requested earlier.
Whatever we had in the 2017.09.27 tag works, so please don't make any changes to the diag_table for the straits at this point!
Meaning we got monthlies in post-processing?
No, dailies. This is what works for the refine scripts that compute mfo:
ncdump -h 00010101.ocean_Denmark_Strait.nc netcdf 00010101.ocean_Denmark_Strait { // format variant: 64bit dimensions: xh_sub01 = 91 ; yh_sub01 = 1 ; z_l = 35 ; z_i = 36 ; time = UNLIMITED ; // (365 currently) nv = 2 ; yq_sub01 = 1 ;
We are saving daily output for the straits data
Sections for CMIP6/OMIP
"ocean_Barents_opening", 1, "days", 1, "days", "time" "ocean_Bering_Strait", 1, "days", 1, "days", "time" "ocean_Davis_Strait", 1, "days", 1, "days", "time" "ocean_Windward_Passage", 1, "days", 1, "days", "time" "ocean_Denmark_Strait", 1, "days", 1, "days", "time" "ocean_Drake_Passage", 1, "days", 1, "days", "time" "ocean_English_Channel", 1, "days", 1, "days", "time" "ocean_Faroe_Scotland", 1, "days", 1, "days", "time" "ocean_Florida_Bahamas", 1, "days", 1, "days", "time" "ocean_Fram_Strait", 1, "days", 1, "days", "time" "ocean_Iceland_Faroe_V", 1, "days", 1, "days", "time" "ocean_Iceland_Faroe_U", 1, "days", 1, "days", "time" "ocean_Indonesian_Throughflow", 1, "days", 1, "days", "time" "ocean_Mozambique_Channel", 1, "days", 1, "days", "time" "ocean_Pacific_undercurrent", 1, "days", 1, "days", "time" "ocean_Taiwan_Luzon", 1, "days", 1, "days", "time" "ocean_Agulhas_section", 1, "days", 1, "days", "time" "ocean_Gibraltar_Strait", 1, "days", 1, "days", "time" "ocean_Iceland_Norway", 1, "days", 1, "days", "time"
CMIP6 wants months. Hence, we should change the above first instance of "days" to "months".
Sections for CMIP6/OMIP
"ocean_Barents_opening", 1, "months", 1, "days", "time" "ocean_Bering_Strait", 1, "months", 1, "days", "time" "ocean_Davis_Strait", 1, "months", 1, "days", "time" "ocean_Windward_Passage", 1, "months", 1, "days", "time" "ocean_Denmark_Strait", 1, "months", 1, "days", "time" "ocean_Drake_Passage", 1, "months", 1, "days", "time" "ocean_English_Channel", 1, "months", 1, "days", "time" "ocean_Faroe_Scotland", 1, "months", 1, "days", "time" "ocean_Florida_Bahamas", 1, "months", 1, "days", "time" "ocean_Fram_Strait", 1, "months", 1, "days", "time" "ocean_Iceland_Faroe_V", 1, "months", 1, "days", "time" "ocean_Iceland_Faroe_U", 1, "months", 1, "days", "time" "ocean_Indonesian_Throughflow", 1, "months", 1, "days", "time" "ocean_Mozambique_Channel", 1, "months", 1, "days", "time" "ocean_Pacific_undercurrent", 1, "months", 1, "days", "time" "ocean_Taiwan_Luzon", 1, "months", 1, "days", "time" "ocean_Agulhas_section", 1, "months", 1, "days", "time" "ocean_Gibraltar_Strait", 1, "months", 1, "days", "time" "ocean_Iceland_Norway", 1, "months", 1, "days", "time"
Otherwise, we need to write postprocessing code to properly handle the monthly averages, which seems unnecessary and will take more debugging time than we have now. I see no reason to keep daily output when CMIP6 wants months.
Andrew Shao is writing the refineDiag script assuming we will have monthly output. Please make the change to the diag table accordingly.