Closed penguian closed 4 years ago
@rml599@nci.org.au commented
Also need to be aware that ktau behaves differently between online atmosphere-only and online coupled runs. In atmosphere-only ktau continues to increment across restart, in coupled run ktau returns to zero at each restart.
@bep599@nci.org.au commented
To prevent the initialization codes affecting offline runs, the block should be moved to the online interface as soon as possible. In the meantime, a quick fix suggestion: IF( .NOT.cable_user%cable_runtime_coupled .and. .NOT.cable_runtime%offline) THEN Or have a new flag to represent the actual usage of the block: IF( cable_runtime%amip ) THEN
@jxs599@nci.org.au changed milestone from 6. Report
to 4. Low Priority
@jxs599@nci.org.au changed status from new
to closed
@jxs599@nci.org.au set resolution to fixed
@jxs599@nci.org.au changed milestone from 4. Low Priority
to 1. Closed
@jxs599@nci.org.au commented
pushed to trunk @ 2140513551ef2be42bcfc8ebf321b1e73853c4a7
@ccc561@nci.org.au changed keywords from wb, wbice reinitialized
to wb, wbice reinitialized nogit
keyword_wb
keyword_wbice_reinitialized_nogit
owner:ewk599@nci.org.au
resolution_fixed
type_defect
| by bep599Need to rethink the initialization put in place within the soil_snow module, preferably moved to the online interface and offline driver. Currently, these are happening for ktau=1 (local variable, which should also use another name to prevent confusion) for some variables and ktau_gl=1 for others. As the variable ssnow%wb and wbice are revised at the first time step, this strategy poses a barrier to data assimilation in the offline runs (not sure about online). In addition, it messes up the water balance for the first time step (which could be why subroutine mass_balance is discarding the numbers from the first 10 time steps). This strategy aims at keeping the soil water moisture within reasonable bounds at the start of simulation. If that concern is paramount, the values of wb and wbice should be checked for each time step; but at present, their values can drift outside those boundaries during a continuous run.
Issue migrated from trac:71 at 2023-11-27 11:11:51 +1100