Update gw_convect #6749
Update gw_convect #6749
Conversation
|
whannah1
force-pushed
the
whannah/update-gw-convect-src
branch
from
dbc44a1 to
5c37d1b
Compare
|
|
||
| subroutine gw_convect_init( plev_src_wind, mfcc_in, errstring) | ||
| use ref_pres, only: pref_edge | ||
| real(r8), intent(in) :: plev_src_wind ! previously hardcoded to 70000._r8 |
There was a problem hiding this comment.
I'd add here in the comments that plev_src_wind should be in units Pa to match pref_edge. Maybe also want to mention that actual steering flow pressure level can differ from what the user requests wherever the reference pressure values differ from the actual pressure values (i.e. over mountainous terrain).
| <entry id="gw_convect_hdepth_min" type="real" category="gw_drag" | ||
| group="gw_drag_nl" valid_values="" > | ||
| minimum hdepth for for convective GWD spectrum lookup table | ||
| Default: 2.5 |
| real(r8), parameter :: AL = 1.0e5_r8 | ||
| ! fixed parameters (we may want to expose these in the namelist for tuning) | ||
| real(r8), parameter :: tau_avg_length = 1.0e5_r8 ! spectrum averaging length | ||
| real(r8), parameter :: heating_altitude_max = 20e3 ! max altitude to check heating (probably don't need this) |
There was a problem hiding this comment.
Add units to heating_altitude_max description. Also, 20000._r8?
There was a problem hiding this comment.
fixed - the change in how this value is specified won't cause a BFB change.
| @@ -118,6 +118,11 @@ module gw_drag | |||
| ! namelist | |||
| logical :: history_amwg ! output the variables used by the AMWG diag package | |||
|
|
|||
| logical :: use_gw_convect_old ! switch to enable legacy behavior | |||
| real(r8) :: gw_convect_plev_src_wind ! reference pressure level for source wind for convective GWD | |||
There was a problem hiding this comment.
Add that gw_convect_plev_src_wind must be in units Pa.
There was a problem hiding this comment.
@whannah1 , do we have simulations that documents the behavior with the new Beres scheme for heating depth etc. (i.e., use_gw_convect_old = .false.).
| @@ -118,6 +118,11 @@ module gw_drag | |||
| ! namelist | |||
| logical :: history_amwg ! output the variables used by the AMWG diag package | |||
|
|
|||
| logical :: use_gw_convect_old ! switch to enable legacy behavior | |||
| real(r8) :: gw_convect_plev_src_wind ! reference pressure level for source wind for convective GWD | |||
| real(r8) :: gw_convect_hdepth_min ! minimum hdepth for for convective GWD spectrum lookup table | |||
There was a problem hiding this comment.
Add units for gw_convect_hdepth_min.
| logical :: use_gw_convect_old ! switch to enable legacy behavior | ||
| real(r8) :: gw_convect_plev_src_wind ! reference pressure level for source wind for convective GWD | ||
| real(r8) :: gw_convect_hdepth_min ! minimum hdepth for for convective GWD spectrum lookup table | ||
| real(r8) :: gw_convect_storm_speed_min ! minimum convective storm speed for convective GWD |
There was a problem hiding this comment.
Add expected units for gw_convect_storm_speed_min.
|
e3sm_atm_developer passes on Chrysalis, but it didn't do the baseline comparison (my mistake). I ran a single "SMS.ne4pg2_oQU480.F2010.chrysalis_intel.eam-cosplite" test with baseline comparison enabled and it passes - so my expectation of these changes being BFB holds up. |
Update gw_convect The PR makes several changes to the convective gravity wave methods. The primary motivation was to address a problem in gw_beres_src() where the diagnosed heating max/depth values were erroneously too low whenever the ZM heating profile included negative values (or zero) in the middle of the heating layer. Additionally, previously hard coded values were exposed as namelist values for a planned QBO tuning exercise for the QBO SciDAC project. [BFB] [NML] new nml variables for eam tests with default physics.
|
Merged to next. |
The PR makes several changes to the convective gravity wave methods. The primary motivation was to address a problem in gw_beres_src() where the diagnosed heating max/depth values were erroneously too low whenever the ZM heating profile included negative values (or zero) in the middle of the heating layer. Additionally, previously hard coded values were exposed as namelist values for a planned QBO tuning exercise for the QBO SciDAC project.
[BFB]
[NML] new nml variables for eam tests with default physics.