6.18. science_fixes.nml
¶
This file contains one namelist called JULES_TEMP_FIXES
.
This namelist sets ‘short-term’ temporary logicals used to protect science bug fixes that lead to alterations in science results. It is expected that these logicals will be short lived as the preference should be for all configurations to use the corrected code. However, to maintain short term reproducibility of results across JULES versions the fixes are protected by logicals until the fixes become the default in all model configurations at which point the logical is retired. See module for when the switch is due for review.
6.18.1. JULES_TEMP_FIXES
namelist members¶
-
JULES_TEMP_FIXES::
ctile_orog_fix
¶ Type: integer Permitted: 0-2 Default: 2 If nonzero, corrects the surface exchange calculations in coastally tiled grid-boxes, assuming that the lowest level is physically terrain following and adjusting the temperature of the land/sea portions in accordance with their relative offset from the grid-box mean height using a dry/moist lapse rate where appropriate. Option 2 will only adjust values over the sea.
-
JULES_TEMP_FIXES::
l_accurate_rho
¶ Type: logical Default: F This switch improves the calculation of surface air density in the surface turbulent fluxes. It includes appropriate use of dry air density when the atmospheric water vapour is expressed as a mixing ratio (l_mr_physics = .TRUE.), otherwise use the wet air density when it is expressed as a specific humidity.
-
JULES_TEMP_FIXES::
l_dtcanfix
¶ Type: logical Default: F This switch corrects a bug in the evolution of the skin temperature in the implicit solver, whereby the change in the skin temperature is artificially constrained. This generally has a small effect, but can cause unphysical results if a canopy with a large heat capacity is coupled to an underlying substrate with a small heat capacity.
-
JULES_TEMP_FIXES::
l_fix_alb_ice_thick
¶ Type: logical Default: F When zero-layer sea ice is used the thermodynamics is calculated in the UM through an effective thickness calculated from snow and ice thicknesses and associated thermal conductivities. With multi-layer sea ice the thermodynamics is calculated in the sea ice component of the model, and the effective thickness is no longer required. However, it was still being used erroneously. This fix removes the effective thickness adjustment when multi-layer sea ice is used.
-
JULES_TEMP_FIXES::
l_fix_albsnow_ts
¶ Type: logical Default: F The original version of the two-stream scheme to calculate the albedo of snow in JULES contains a bug in the calculation of the reflection coefficient that renders very thin layers of snow too reflective. This logical applies the appropriate correction when it is enabled.
-
JULES_TEMP_FIXES::
l_fix_lake_ice_temperatures
¶ Type: logical Default: F If true, allows sea ice temperatures in lakes to evolve over time for coupled models when the lake is defined as a sea point but is not coupled to an ocean model.
-
JULES_TEMP_FIXES::
l_fix_moruses_roof_rad_coupling
¶ Type: logical Default: F If true, this switch corrects a bug in the surface energy balance when the MORUSES radiative roof coupling is used (see
l_moruses_storage
). If false, the thermal conductivity of the soil (hcons) is erroneously set to zero, which causes the roof to be effectively uncoupled whenl_vegcan_soilfx
.
-
JULES_TEMP_FIXES::
l_fix_osa_chloro
¶ Type: logical Default: F When set to false, the chlorophyll content used to determine the optical properties of water, for the ocean surface albedo, are specified in gm-3 when the parameterisation they use is defined in mg m-3. It is a short term logical until the code becomes the new default.
-
JULES_TEMP_FIXES::
l_fix_snow_frac
¶ Type: logical Default: F When set to false, there is the potential to have small snow mass, but a zero snow fraction due to machine precision in the calculations. This prevents sublimation or snow melt from removing the remaining snow mass, hence small values can persist. In addition to this there is a conceptual bug in the calculation of the fraction of potential evaporation because it does not add in canopy evaporation when the snow fraction is less than one. When set to true these issues are corrected and in addition the radiation calculations for snow fraction are also made consistent.
-
JULES_TEMP_FIXES::
l_fix_ustar_dust
¶ Type: logical Default: F If true, corrects how ustar is calculated in the exchange coefficient for dust deposition
-
JULES_TEMP_FIXES::
l_fix_wind_snow
¶ Type: logical Default: F If true, ensures that wind speed is calculated for use in snow unloading. If false, the wind speed for unloading will be zero on timesteps when 10m wind diagnostics are not calculated. This will tend to leave more snow on the vegetation. It is a short term logical until the code becomes the new default.