Missing "SpecificStorage.FileName" key

[Juan Sebastian Acero Triana]

Good afternoon,

I was looking for the key to define specific storage using a pfb file in the manual and the code and did not find any. Among all the subsurface parameters, specific storage is the only one missing this key in the code. Is there a reasonable reason for this?

I also notice the description of the "Porosity.FileName" key is missing in the manual but present in "input_porosity.c". Additionally, the unique type for Porosity and SpecificStorage reported in the manual is Constant (i.e., no PFBfile). If this is not the case, the manual should be updated.

Kind regards,

Juan S. Acero Triana

[Chen Yang]

Hey Juan,

Thanks for your question.

You can use the indicator file to label each geologic/soil unit and assign a parameter group (including porosity, permeability, VG parameters, specific storage etc.) to each indicator. This has the same effects as you use the pfb file. Since the indicator file can handle any heterogeneity of the subsurface materials, PFB for each individual parameter is not a necessary choice. And Indicator file in fact is much more convenient to manage your model since the same geologic/soil unit always has the same parameter group. This avoids errors if you prepare pfb files separately for each parameter.

Thanks,

Chen

 

-- 

Chen Yang (杨晨), Ph.D.

Associate Research Scholar

Department of Civil and Environmental Engineering

Princeton University

 

From: <par...@googlegroups.com> on behalf of Juan Sebastian Acero Triana <juan...@ucr.edu>
Date: Monday, May 9, 2022 at 5:50 PM
To: ParFlow <par...@googlegroups.com>
Subject: Missing "SpecificStorage.FileName" key

 

Some people who received this message don't often get email from juan...@ucr.edu. Learn why this is important

--
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 on the web visit https://groups.google.com/d/msgid/parflow/1c0628d0-66a7-440f-b3d0-f34140827254n%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Juan Sebastian Acero Triana]

Hi Chen,

Thank you for your prompt reply. I may have parameters with more than 1,000 unique values, and several may not be specific to a single geologic unit (no same parameter group). So I do not find the indicator file approach appealing, as I would need to assign values to many indicators in the TCL file. Could you please confirm if all parameter values can be read directly from pfb files? If yes, I might only need to use the indicator file approach for the specific storage as its FileName key is missing in the code.

Thank you,

Juan

--

Juan S. Acero Triana, PhD

Postdoctoral Fellow

Department of Environmental Sciences

University of California, Riverside

Room 2319 Geology
Riverside, California 92521
Email: juan...@ucr.edu

[Chen Yang]

Hi Juan,

There is no PFBFile input for specific storage and there is PFBFile input for porosity.

We always set specific storage as a constant probably due to the sensitivity or data source.

Thanks,

Chen

 

-- 

Chen Yang (杨晨), Ph.D.

Associate Research Scholar

Department of Civil and Environmental Engineering

Princeton University

 

From: Juan Sebastian Acero Triana <juan...@ucr.edu>
Date: Monday, May 9, 2022 at 8:38 PM
To: Chen Yang <chen...@princeton.edu>
Cc: ParFlow <par...@googlegroups.com>
Subject: Re: Missing "SpecificStorage.FileName" key

 

You don't often get email from juan...@ucr.edu. Learn why this is important

[Engdahl, Nick]

Hi Juan,

Chen’s assessment that ParFlow does not currently have an option to do what you’re trying to set via keys is correct but the implication that this was a philosophical decision by the development team is not. We always try to let the user decide what is best but a few things have slipped through the cracks; you happen to have found one. I completely agree with you that indicator files and solid files are not always the best choice but a few parameters, like this, still require them for now.

 

So, how to address your issue? I’d look for zoning/clustering of your storage data to define regions where an average would be a reasonable approximation and define those zones as geometries. The most general option to do this is with a solid file since those are grid-independent, but indicators can work too. There are also ways of partially accounting for some effects of spatial variability in storage relative to a mean storage value by scaling the permeability values, though those doing so comes with some other caveats. The argument here is that usually there is joint uncertainty in storage and permeability when the data are reported, so that uncertainty can be “shifted around” in the model via effective parameter values. This gets to what I think Chen was alluding to, but let’s be clear that specific storage is not required to be set as a constant, even if it often is for reasons of confidence/uncertainty. Beyond that, the tradeoffs involved in different solutions get very philosophical very quickly, so I’d suggest attending the next ParFlow Short Course (and Modflow & More 2022) to dig deeper into these issues with us.

 

Nick

 

------------------------------------------------------------

Nicholas B. Engdahl, PhD

Associate Professor of Civil and Environmental Engineering

Washington State University

PO Box 645815

Pullman, WA 99164-5815

https://labs.wsu.edu/groundwater/

------------------------------------------------------------

To view this discussion on the web visit https://groups.google.com/d/msgid/parflow/9DB19ED8-1D74-4876-B260-53609C7F12BF%40princeton.edu.