…phase-2-update

use existing lnd_domain attribute in infodata
we do not differentiate on coupler side if data land or not
still need to send the proper land domain file

domain file is actually set, in shr_strdata_type%domainFile
all the other data models use the first stream file for domain
information

land is the one that actually has that member set in the SDLND
structure

it is easier to get it from there, then from stream
the domain file in regular land case is defined by fatmlndfrac file

iMOAB_ApplyScalarProjectionWeights is not used
anymore
all maps are done now with seq_map method, and there is only
one instance of iMOAB_ApplyScalarProjectionWeights in E3SM
using moab driver

only intel on chrysalis complained about it
It could be a cause for trouble later on

…y defaults

* Disable timing in prim_main for the first two teps
* Add some timers in EulerStepFunctor

Also fix preqx's use of use_moisture. Remove MOIST-DRY enum to avoid confusion,
since it's no longer used.

compute rof lnd interaction only if rof_c2_lnd is true
we were computing an unnecessary map always
even if rof_c2_lnd is false
(for most coupled cases it is true)

in case rof c2 lnd, we were not retrieving the the MPI group
needed for comm graph computation, for same grid path
we were retrieving it for intx only
how come we never had samegrid land rof before ?

…ing and continue (instead of failing) when input has month dimension and output does not.

… modern compilers (per Xylar)

… redundant

This adds compsets and new output variables to facilitate the E3SM and E3SM-MMF
contributions to phase 1 & 2 of the radiative-convective equilibrium model inter-comparison project (RCEMIP).

NOTE: in the original RCE compsets there was a slight error in the value of constant_zenith_deg being off by 1/100. This difference is inconsequential, but I wanted to fix it here to ensure consistency with the other RCEMIP models.

[non-BFB] - only for RCE cases because of updated constant_zenith_deg value

…ject#6621)

harden eamxx/prod testmod

makes the ERS eamxx/prod test compare outputs in all files, requiring pulling a bugfix from the scream repo.

[BFB]

Signed-off-by: Jason Boutte <[email protected]>

Changes needed for spot/aurora: sycl kokkos spaces,
temporary TP changes (TP init will be sorted out later),
switch booleans to ints in F<-->xx interfaces.

see E3SM-Project#6569 for the original branch/history.

[bfb]

Add data land capability for moab driver;
case tested: --res r05_r05 --compset RMOSGPCC
standard data moab model, mainly adapt from data ocean implementation

major change: Initialize all factors set by MOAB area correction init routine to 1.0. It affects all models,
but it seems that some values were not initialized for data land case.
On chrysalis, with intel compiler, it actually complained about non initialized values when applying correction factors.

The compressing file errors seen during development were caused by non-ascii characters in the log files.
Those were caused by an uninitialized mapper, for which we were still creating a MOAB map. It was never used,
it is fixed properly now, the moab map is not computed anymore.

[BFB]

MALI update to fix issues from earlier PR causing sporadic test failures

Including a variable that was deactivated in the globalStats stream
caused sporadic failures during the second run of some ERS tests on
several platform/compiler combinations. That variable is now only
included when MALI is using Albany. Also updates a namelist default
that had been missed but does not change answers.

Fixes E3SM-Project#6600

[NML] for configurations with MALI
[BFB]

…3SM-Project#6633)

* Adds dane/ruby queue definitions
* Updates dane/ruby cmake macros
* Updates dane/ruby machine definition to use slurm and new modules
* Removes lc_slurm

…Project#6614)

This new feature allows `interpinic` to interpolate an ELM initial conditions file
that require storage in `NC_FORMAT_64BIT_DATA` and other netCDF4 formats.

[BFB] [Bugfix]