Perturbing SMOIS in rst file

[nef]

Dear all,

I am using WRFV4.6 with the NoahMP LSM. I want to perturb the top layer soil moisture of a restart file (wrfrst) and continue the simulation. However when I try to restart it the model crashes because the water balance is off (ERRWAT>0.1). Is there a way to circumvent that?

Thank you in advance,
Nefeli

 

[Ming Chen]

Nefeli,
If you manually change the value of a physical variable, you need to be sure the changes are physically reasonable. Otherwise, the model will crash as you have seen. I would suggest you start from a minor change, and see how the model responds to your changes. There is no rule for us to follow about what changes are reasonable or not. So such tests are purely empirical.

 

[nef]

Thank you so much for your quick response! It seems, however, that the model cannot accept even very small variations in volumetric soil moisture (of the order of 0.01 m^3/m^3), bacause it checks the water balance in every grid point and crashes if the water balance is altered more than what can be attributed to numerical diffusion. Since my research focuses specifically on perturbing the surface soil moisture, would it be unwise to code out the water balance check, and what considerations should I take into account if I do so?
Thank you for your help!

 

[vaakos]

Hi Nefeli,

We were able to circumvent this issue with a simple modification to the module_sf_noahmplsm.F code.

After the water error (ERRWAT) is calculated as follows:

if (calculate_soil) then
ERRWAT = ACC_DWATER - (ACC_PRCP + IRFIRATE*1000.0 + IRMIRATE*1000.0 - ACC_ECAN - &
ACC_ETRAN - ACC_EDIR - RUNSRF - RUNSUB - QTLDRN)

we added the following line:

ERRWAT = 0

This modification prevents the error message from appearing.

However, I'm not sure whether this introduces any significant inconsistency in the model.

Best,
Akos

 

[nef]

Thanks Akos, I did something similar! Please let me know if you notice any significant inconsistencies in the output, I have not noticed anything so far.

Nefeli

 

[vaakos]

Our 6-month continuous simulations with the modified soil moisture look fine; nothing suspicious or unusual.

 

[Ming Chen]

Hi Nefeli,

I don't think it is a good idea to comment out those codes for water balance check, ---- this is an important physical condition that the model results must maintain. By commenting out this check, and if the water balance is not maintained, physics may go wrong and eventually leads to model crash.

This is my personal opinion. Let's see whether someone in the community may propose new ideas.

Thank you so much for your quick response! It seems, however, that the model cannot accept even very small variations in volumetric soil moisture (of the order of 0.01 m^3/m^3), bacause it checks the water balance in every grid point and crashes if the water balance is altered more than what can be attributed to numerical diffusion. Since my research focuses specifically on perturbing the surface soil moisture, would it be unwise to code out the water balance check, and what considerations should I take into account if I do so?
Thank you for your help!