Injecting moisture at a specific grid point throughout a WRF run

[vaakos]

Hello,

I would like to investigate the effects of continuously injecting moisture into the atmosphere at the surface at a specific grid point during a multi-day mesoscale WRF simulation. The moisture should be added throughout the entire run (e.g., at every time step or hourly), not just at initialization. It is also important that the added moisture is fully incorporated into the model's dynamics and physics.

Initially, I considered simply increasing the latent heat flux (LH), but I’m not sure if that would achieve the desired effect.

Are there any guidelines or recommended methods for implementing this? Any advice or experience with similar modifications would be greatly appreciated.

Many thanks in advance,
Akos

 

[vaakos]

As an update: I might have found a way to do this. I believe I should include the moisture source in the specific grid cell by adding it to QFX (the upward moisture flux at the surface, in kg/m^2/s).

Since I'm using the Unified Noah LSM (WRF v4.7.0), I need to modify the module_sf_noahdrv.F file.

Specifically, after line 1259:

QFX(I,J) = ETA_KINEMATIC

I would add something like:

IF (I == isource .AND. J == jsource) THEN
QFX(I,J) = QFX(I,J) + qfx_add
END IF

However, it's unclear whether I need to insert this modification after every occurrence of the line QFX(I,J) = ETA_KINEMATIC, or if it's sufficient to do so only after the first (since this line appears multiple times in the code).

Any expert insights or confirmation would be greatly appreciated.

Akos

 

[kwerner]

Hi, and apologies for the long delay in response while our team tended to time-sensitive obligations. Thank you for your patience.

It looks like the second two instances of that coded line are within the subroutine for the mosaic subtiling option. If you're not using that, I assume you probably don't need to modify those two lines, so just the first should be sufficient. You can always play around with the code and test it out to see how it affects the run.

 

[vaakos]

Thank you very much for your response. It seems to work as I described above.

 

[vaakos]

Hi again,

I'm currently testing the simulations with enhanced QFX at a specific grid point, as we discussed earlier.
Everything looks reasonable during the first few hours. QVAPOR increases near the injection point, as shown in Figure 1.

After a few more hours, though, I start noticing what looks like random noise in the difference fields, appearing far from the moisture source (see Figures 2 and 3). I do not think this is directly caused by the modified QFX, but I am not entirely sure.

Is this kind of behavior expected? And if so, is there any way to reduce or avoid this noise?
It makes it harder to clearly interpret the meteorological response.

Thanks a lot in advance for your help.

Best,
Akos

 

[kwerner]

Hi,
This is likely somewhat of a "butterfly effect." Even small perturbations can grow fast in convective areas. The areas with noise are likely convection locations that have slightly changed but would still generally be the same. Model differences can move at the speed of sound but on the time scales shown it could also be advective. If you chose a period with less convection, you would see less of this. So I'm not sure there's a way to remove it, unfortunately.