raise Exception('!!!!! ERROR: SRTM data not available for regions above 60 deg. Max:'+str(North)+' Min:'+str(South))
Exception: !!!!! ERROR: SRTM data not available for regions above 60 deg. Max:2 Min:-82
_ancil_soils_hydr failed due to running out of time
I was just wondering if there is some way to extend my domain beyond 60 S? Given the SRTM data not avalable beyond 60S, could we use another source to generate _ancil_mean_orog?
Thanks!
P.S. The domain I would like to simulate
For [-40, 99] with [0.1, 0.1] and [800, 1200]:
coords = 39.000, 159.000, -80.000, 0.000
P.P.S. For my simulation, I actually do not need very accurate orography data around the Antarctic. Would it be possible/reasonable to just mask out the orography flat beyond 60 S?
Hi @chenhui.jin, the SRTM dataset for orography only extends to -60. You can though use the GLOBE dataset instead. This is set in the regional ancillary suite. There is a true/false toggle for {}_{}_orog_srtm under the resolution setup.
Please be aware though that BARRA does not extend even to -60 so you will need to use either era5-land for the land/surface or era5 (I would use era5-land, and shift the lat/lons to a grid-corresponding to era5-land points with 0.1 degree spacing).
NOTE: You would need to switch both rg01_rs01_orog_srtm and rg01_rs02_orog_srtm to .FALSE. .
There are still some issues with _ancil_soils_hydr (due to running out of time), and _ancil_cap_sstice (job.err doesn’t show anything helpful to find where the issue occurred).
Using the cylc session cylc.cj0591.if69.ps.gadi.nci.org.au
Loading cylc7/24.03
Loading requirement: mosrs-setup/2.0.1
gc_abort (Processor 0): ANCIL EXIT
--------------------------------------------------------------------------
MPI_ABORT was invoked on rank 0 in communicator MPI_COMM_WORLD
with errorcode 9.
NOTE: invoking MPI_ABORT causes Open MPI to kill all MPI processes.
You may or may not see output from other processes, depending on
exactly when Open MPI kills them.
--------------------------------------------------------------------------
[FAIL] rose mpi-launch -v central_ancillary.exe # return-code=9
2026-03-25T11:08:19Z CRITICAL - failed/EXIT
Hi @chenhui.jin I was talking to @alanah.chapman yesterday at the roundtable. Alanah is going to try to run rAM3 at very high latitudes for some sea ice / aerosol studies.
We were discussing the same issues you are raising, and we discovered this old thread on the hive.
I would reach out to @ZhangchengPei to ask about using the other orography datasets that have coverage over Antarctica.
I don’t believe anyone has yet run rAM3 using those other datasets. But there is definite interest in doing so.
Hi @chenhui.jin, I suggest that you reach out to @ZhangchengPei because it might be a dataset issue near Antarctica (given it is a SST issue). Otherwise I would suggest shifting your lower boundary northward.
Chenhui - if you want to run the model with unrealistic southern boundary conditions, you could try and stay with the srtm orogoraphy and do what Chermelle has suggested above.
If you comment out lines 168-169 of /src/contrib/Apps/Orography/concant_srtm_tiles.py
if North> 60 or South < -60:
raise Exception('!!!!! ERROR: SRTM data not available for regions above 60 deg. Max:'+str(North)+' Min:'+str(South))
You can generate SRTM orography over Australia and flat nothing all the way to the South Pole.
You will need to ensure your land sea mask is also consistent (no land below 60 south), otherwise you will generate soil and vegetation ancils in this region.
I have an idea how you can do this if you want to pursue this (i.e. generation ancils using SRTM but enforce no land south of 60 degrees).
@cbengel’s method: switch off {}_{}_orog_srtm, while we will still have some issues about _ancil_cap_sstice, which I will reach out to @ZhangchengPei for a solution.
Alternatively, I switched _sstsmc_anc to False, and all tasks are completed, and I assume I could run the model without _ancil_cap_sstice for my application?
Paul’s method: For my understanding, we still generate the ancils from STRM, but manually mask land flat south of 60 S?
For [-42, 81] with [0.1, 0.1] and [820, 1600]:
coords = 1.000, 161.000, -83.000, -1.000
For [-42, 81] with [0.1, 0.1] and [800, 1580]:
coords = 2.000, 160.000, -82.000, -2.000
I used Paul’s ancillary checker to check the ancillaries. There are NaNs around the Antarctic, mainly associated with soils
The first task in the regional ancillary suite : <regional>_<resolution>_ancil_lct generates a land-sea mask that is used to define all downstream soil and vegetation fields.
Here is what the cylc graph for the suite looks like:
So I believe you can use this suite to remove all soil and vegetation ancillary data. What I would do is
Run the regional ancillary suite using your current domain definitions, but begin the suite with rose suite-run -- --hold. This will load the suite, but hold all tasks.
Manually trigger the upstream tasks install_cold_idl, install_cold_hpc, install_ants & install_ugants and <Region>_<resolution>_ancil_top. Do this by right-clicking on the task-name and selecting ‘Trigger now’ in the pop-up menu.
Manually trigger all the <Region>_<resolution>_ancil_lct tasks you have.
These will create the files qrparm.mask which is the land-sea mask used by each resolution (or nest).
Manually edit this file to remove all land points south of 60 degrees. You can do this by using the iris module and the iris.load_cube() method. Scott Wales wrote an example of this on the old CLEX wiki - I’m hosting it now on the 21st Century Weather wiki here:
Once you have saved your updated qrparm.mask file with all land points removed south of 60 degrees, you can ‘release’ all the downstream tasks in your suite. Use the cylc GUI pull-down menu from the top taskbar : ‘Control’ → ‘Release Suite (unpause)’.
This should generate a set of land, soil and vegetation ancillaries that are consistent with a semi-aquaplanet (i.e. no Antarctica).
BUT this may only be the first step in creating this make-believe scenario.
You may have to do some edits with
a) Your sea-surface temperature datasets, as they are probably affected by real observed sea ice
b) Your snow and ice related climatology ancillary files (although it sounds like you are already doing this).
It’s an interesting problem, but that’s how I’d go about generating a semi-aquaplanet experiment (if this is what you’re after).
