Closed edoddridge closed 6 years ago
Merging #196 into master will decrease coverage by
0.47%
. The diff coverage is94.41%
.
@@ Coverage Diff @@
## master #196 +/- ##
=========================================
- Coverage 90.98% 90.5% -0.48%
=========================================
Files 19 19
Lines 1409 1570 +161
Branches 76 82 +6
=========================================
+ Hits 1282 1421 +139
- Misses 107 123 +16
- Partials 20 26 +6
Impacted Files | Coverage Δ | |
---|---|---|
aronnax/driver.py | 93.69% <ø> (ø) |
:arrow_up: |
src/thickness.f90 | 97.64% <100%> (+0.05%) |
:arrow_up: |
src/aronnax.f90 | 90% <100%> (+0.1%) |
:arrow_up: |
src/state_deriv.f90 | 91.66% <100%> (ø) |
:arrow_up: |
src/model_main.f90 | 95.65% <100%> (+1.68%) |
:arrow_up: |
src/time_stepping.f90 | 58.49% <100%> (ø) |
:arrow_up: |
src/momentum.f90 | 92.1% <93.87%> (-7.9%) |
:arrow_down: |
src/enforce_thickness.f90 | 36.36% <0%> (-27.28%) |
:arrow_down: |
Continue to review full report at Codecov.
Legend - Click here to learn more
Δ = absolute <relative> (impact)
,ø = not affected
,? = missing data
Powered by Codecov. Last update c86071e...8f346fa. Read the comment docs.
Should have mentioned that the subroutine for enforcing layer thickness remains greater than hmin
has been left in for now, but should probably be deleted.
This PR implements a substantial new feature, namely outcropping and grounding of isopycnals, thereby closing #26.
There is a new parameter
wind_depth
that controls how deep the wind forcing is felt - the forcing is spread evenly over this depth, and any layer withinwind_depth
metres of the surface will receive a proportion of the forcing. The old behaviour of putting all the forcing into the top layer is preserved ifwind_depth
is set to zero, or left unset, in which case it defaults to zero.The way the model implements
hmin
has been drastically changed - it no longer forces layers to be thicker thanhmin
. Instead, the thickness diffusivity is ramped up exponentially as the layer thickness approacheshmin
, which creates thin, flat layers. The old behaviour of simply inflating layers until they were as thick ashmin
was an ugly hack that destroyed any hope of volume conservation, and it has not been preserved.A new test has been added. It is a reentrant channel with two layers and sloping bathymetry (which helps with the checklist in #6), and each of the layers has an initial thickness of 0.1 m somewhere in the domain.