opposite direction of ground heat flux in urban models with Noah-MP land surface model

The ground heat flux in Noah-MP is downwards positive from surface, but it is downwards negative from surface in Noah land surface model (LSM).

To keep the same direction of ground heat flux between urban models (SLUCM, BEP, or BEM) and the land surface model (Noah or Noah-MP) is required to calculate surface energy balance at the grid mean level.

In Noah LSM, the ground heat flux direction is the same as urban models, but differ in Noah-MP.

To make the same direction using SLUCM with Noah-MP LSM at the grid mean level.

hrldas/urban/wrf/NoahmpUrbanDriverMainMod.F

Line 530 in f550db9

    
           GRDFLX(I,J)   = FRC_URB2D(I,J) * (G_URB) + (1-FRC_URB2D(I,J)) * GRDFLX(I,J)        ![W/m/m]

should revise to
GRDFLX(I,J) = FRC_URB2D(I,J) * (-1.0 * G_URB) + (1-FRC_URB2D(I,J)) * GRDFLX(I,J) ![W/m/m]

and for BEP, BEM

hrldas/urban/wrf/NoahmpUrbanDriverMainMod.F

Line 759 in f550db9

    
           grdflx(i,j)   = (1.-frc_urb2d(i,j))*grdflx_rural(i,j)+ frc_urb2d(i,j)*grdflx_urb(i,j)

should revise to
grdflx(i,j) = (1.-frc_urb2d(i,j)) * grdflx_rural(i,j)+ frc_urb2d(i,j) * grdflx_urb(i,j) * -1.0

@tslin2 Could you please double check BEP and BEP-BEM? I remembered that in these two models, the sign is correct? Maybe I remembered it wrong.

@cenlinhe
my simulation results show that in Noah with urban models, the GRDFLX has the same direction in urban and rural areas, but is opposite between rural and urban areas in Noah-MP with urban models including using SLUCM, BEP, or BEP-BEM.

OK, we need to fix this then in both hrldas and wrf. You may also want to open a WRF issue to point out this issue.

I confirmed this bug.
The ground heat flux is positive for downward direction in Noah-MP: https://github.com/NCAR/noahmp/blob/master/src/SurfaceEnergyFluxBareGroundMod.F90#L155
the ground heat flux is negative for downward direction in SLUCM: https://github.com/NCAR/hrldas/blob/master/urban/wrf/module_sf_urban.F#L1503
and BEP: https://github.com/NCAR/hrldas/blob/master/urban/wrf/module_sf_bep.F#L3287
and BEP_BEM: https://github.com/NCAR/hrldas/blob/master/urban/wrf/module_sf_bep_bem.F#L5312

This bug has been fixed in this commit for v5.0: a3ce25e
and in this commit for v4.5: NCAR/noahmp@a26f91f
Here I am closing this issue.