About the result of saturation in normal simulation

[Zeyu Tang]

Dear everyone, 

     I want to ask a troubled question, When I turn on the overlandflow and end the second spinup phase into the normal simulation,  I have observed that the saturation values in the formal simulation are predominantly close to 1 (the red color).

     Could you kindly provide any insights or suggestions on why this might be occurring and how it could be addressed?

[Reed M. Maxwell]

Hi- It’s very hard to answer your question without more information.  If this is just seepage face BC to overland flow, with a well-conditioned and processed DEM the differences are somewhat minimal.  If you can provide more information about your run, domain, how you processed the DEM, what you did for spinup in the first part, your discretization, etc, folks on this list might be able to help you.

 

Reed

 

 

 

     Could you kindly provide any insights or suggestions on why this might be occurring and how it could be addressed?

--
You received this message because you are subscribed to the Google Groups "ParFlow" group.
To unsubscribe from this group and stop receiving emails from it, send an email to parflow+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/parflow/911761b5-850e-4506-afb5-e3eddaade87bn%40googlegroups.com.

[Zeyu Tang]

Thank you for your reply, my answer is as follows:

Model Domain Configuration: My study area is the Danube River Basin. This is a large watershed, I set the model resolution to dx = dy = 10,394.2 m ≈ 10,000 m (0.1°). Could this relatively coarse model resolution be a potential reason for the excessive saturation observed in my simulations?

pfset ComputationalGrid.Lower.X           0.0
pfset ComputationalGrid.Lower.Y           0.0
pfset ComputationalGrid.Lower.Z           0.0

pfset ComputationalGrid.DX                10394.2
pfset ComputationalGrid.DY                10394.2
pfset ComputationalGrid.DZ                200.0

pfset ComputationalGrid.NX                161
pfset ComputationalGrid.NY                82
pfset ComputationalGrid.NZ                10

DEM Processing: I used the terrain processing code from Jun Zhang’s published work, specifically the PriorityFlow algorithm, and validated the computed drainage area against reference values, it is consistent. However, after a parking lot test on the slope, I identified several anomalous ponding grids. I don't know how to address these issues—could they be a source of error in my model?

Spin-up Process: The depth to bedrock (DTB) in the Danube River Basin ranges from 2 m to 48 m, and I set my model domain accordingly with a total depth of 107 m and a layered structure as follow. The spin-up process consisted of two stages:

pfset Cell.0.dzScale.Value                       0.25
pfset Cell.1.dzScale.Value                       0.125
pfset Cell.2.dzScale.Value                       0.05
pfset Cell.3.dzScale.Value                       0.05
pfset Cell.4.dzScale.Value                       0.05
pfset Cell.5.dzScale.Value                       0.025
pfset Cell.6.dzScale.Value                       0.005
pfset Cell.7.dzScale.Value                       0.003
pfset Cell.8.dzScale.Value                       0.0015
pfset Cell.9.dzScale.Value                       0.0005

First stage: The model was driven by long-term average (PME), with an initial groundwater table depth of -2 m. Turn off Overland flow until groundwater storage changes stabilized within 3%.
Second stage: Turn on Overland flow, and the model ran until surface  and subsurface storage, runoff stabilized within 1% of PME.
After spin-up, I repeatedly used 1980 meteorological data to drive the ParFlow-CLM model. A few years later, I analyzed the monthly average saturation distribution for simulation years and found that most areas remained in a fully saturated state.
 

I would like to understand the possible causes of this issue—could it be related to any of the factors mentioned above? And whether I set the total model depth to 392m to be consistent with the CONUS settings or not?  I look forward to receiving everyone's feedback.