jhan_HaC5.5Oct

HAC vn 5.5

Archived Earlier ramblings

Script for mapping/syncing files between HAC and the trunk/ has been organized.

#224 re-organizes trunk into infrastructure format adopted by HAC.

#226 attempts to make all the files the same. The key differences are:

• removal of types
• USEing modules

It was necessary for JAC to organize CABLE-HAC to facilitate calling cable_rad_main on the first timestep (avoided in CMIP 5 & 6). This amounts to

This directly involved radiation and albedo subroutines. In this process these subroutines were re-written. The only "process" change was:

• triggering "sunlit" off (cosine of) zenith angle,

as in principle the zenith angle should give the "same" information and could be determined at this point, where as maintaining the use of the downward shortwave was more problematic (especially when considering porting this to the online UM application)

The proposed trunk (#226) and HAC were compared at Loobos. CABLE-offline Loobos forcing was prepared. Although including N-tiles through the met forcing file was problematic. An adapted version of #199 was used and I think this is OK now. The albedos were very different. Although in the standalone model this didn't really have an impact on the "climate". Nonetheless, investigating this further by plotting quantities leading up to the albedo calculation (and trying to overcome the tiling issue mentioned above) revealed several issues.

1. It would seem that JULES has no use for zenith angle as its values were dropped after the first timestep. Hacking this at least things were comparable.

2. The bare soil albedo was different. Hacking this at least things were comparable.

3. Thins were still at the very least out-of-phase as the cos(zenith) calcs used b/n models are different.