I ran a case in WRFv4.4.2, which was a 15 km run using Noah-MP land surface model.
Simulation for SF_URBAN_PHYSICS = 1, Single-layer UCM in the Southeast United States.
Simulation started from 2015-08-02 00Z and plan to end at 2015-08-10 00Z
The run crashed at 2015-08-03_14:30:00Z due to NaN values for air density and vertical water mixing ratio in the non-urban grid. A huge negative (-1450 W/m2) sensible heat flux was found in an urban grid at 2015-08-02 22Z (see figure), and urban fraction at this grid is around 30%. The default TS_SCHEME is 1 (4-layer model) in the URBPARM.TBL, changing TS_SCHEME from 1 (4-layer model) to 2 (force_restore) in the URBPARM.TBL file can fix this issue and the results look good.
A further diagnosis in the 4-layer model (TS_SCHEME is 1) found that the road temperature dropped too much after rainfall due to increasing latent heat flux in the impervious surface when precipitation is larger than 1mm/hr, and create imbalance of energy budgets. The non-linear energy equation solved by Newton-Rapson method for road and wall temperature is unstable, creating very bad road, urban canopy air, and wall temperature after numerical iteration.
The force restore method (TS_SCHEME is 2) works because it assumes the energy balance by ground heat flux. The ground heat fluxes is residual of net radiation, sensible, and latent flux.
Another solution may assume no latent heat flux in the impervious road after rainfall or changing BETG (minimum moisture availability of road) to 0.1 from 0.7 (rainfall >1mm/hr) in the code. Or assuming the ground heat fluxes is residual of net radiation, sensible, and latent flux before solving energy equation. However, for a quick repair, altering TS_SCHEME in the URBPARM.TBL from 1 (4-layer model) to 2 (force_restore) may be the best temporary option.

Tzu-Shun
Simulation for SF_URBAN_PHYSICS = 1, Single-layer UCM in the Southeast United States.
Simulation started from 2015-08-02 00Z and plan to end at 2015-08-10 00Z
The run crashed at 2015-08-03_14:30:00Z due to NaN values for air density and vertical water mixing ratio in the non-urban grid. A huge negative (-1450 W/m2) sensible heat flux was found in an urban grid at 2015-08-02 22Z (see figure), and urban fraction at this grid is around 30%. The default TS_SCHEME is 1 (4-layer model) in the URBPARM.TBL, changing TS_SCHEME from 1 (4-layer model) to 2 (force_restore) in the URBPARM.TBL file can fix this issue and the results look good.
A further diagnosis in the 4-layer model (TS_SCHEME is 1) found that the road temperature dropped too much after rainfall due to increasing latent heat flux in the impervious surface when precipitation is larger than 1mm/hr, and create imbalance of energy budgets. The non-linear energy equation solved by Newton-Rapson method for road and wall temperature is unstable, creating very bad road, urban canopy air, and wall temperature after numerical iteration.
The force restore method (TS_SCHEME is 2) works because it assumes the energy balance by ground heat flux. The ground heat fluxes is residual of net radiation, sensible, and latent flux.
Another solution may assume no latent heat flux in the impervious road after rainfall or changing BETG (minimum moisture availability of road) to 0.1 from 0.7 (rainfall >1mm/hr) in the code. Or assuming the ground heat fluxes is residual of net radiation, sensible, and latent flux before solving energy equation. However, for a quick repair, altering TS_SCHEME in the URBPARM.TBL from 1 (4-layer model) to 2 (force_restore) may be the best temporary option.

Tzu-Shun
Last edited: