Importing daily climate files into MODFLOW

[miyuru gunathilake]

I would like to know how to import climate data (rainfall) to simulation in MODFLOW through ModelMuse 

Thanks

[Richard B. Winston]

There are several options under "File|Import" that you can use. If your data is not in one of the supported formats, use a GIS to convert it to a supported format. If the projection used in your model differs from the projection of your data, use a GIS to change the projection of your data.

Getting the data into ModelMuse is the easy part. The harder part is to use the data in your model. The problem is that the climate data has only an indirect relationship to the MODFLOW inputs. Precipitation is related to recharge but they are not the same. Recharge is water that reaches the water table. Before precipitation can become recharge, some of it may be intercepted by the canopy. More of it can be lost to surface runoff. Even more can be lost to evapotranspiration from the unsaturated zone. In addition, not all recharge comes directly from precipitation. Some could come from infiltration below lakes and streams.

Temperature and other climate data could be used to estimate potential evapotranspiration. However, the potential evapotranspiration would apply to both the saturated an unsaturated zone whereas MODFLOW requires the potential evapotranspiration from just the saturated zone.

[miyuru gunathilake]

Dear Professor, 

Thanks alot for the answer and your valuable suggestions. 

I will consider your scientific advice. Very useful

--
This group was created in 2004 by Mr. C. P. Kumar, Former Scientist 'G', National Institute of Hydrology, Roorkee. Please visit his webpage at https://www.angelfire.com/nh/cpkumar/
---
You received this message because you are subscribed to a topic in the Google Groups "MODFLOW Users Group" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/modflow/YUSfCDWWvJk/unsubscribe.
To unsubscribe from this group and all its topics, send an email to modflow+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/modflow/6d9fdcbe-384e-4f8a-ad70-823bda5eed4a%40mindspring.com.

[Randall Hanson]

These inputs of precipitation and ET are, in fact, used directly in Modflow-OWHM and then recharge is calculated during the simulation. So this may be a better option for your model.

OWHM2 is also now available in ModelMuse too!! We use OWHM2 regularly for climate change simulations as well as historical climate.

Cheers 

Randy Hanson

One-Water Hydrologic

Sent from my iPhone

On Sep 14, 2023, at 5:26 AM, miyuru gunathilake <miyurubandar...@gmail.com> wrote:



You received this message because you are subscribed to the Google Groups "MODFLOW Users Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to modflow+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/modflow/CAFwE-sjFtB9rs9EJ2VNjsSrpBJdC87HegzLLo3xCDKzE_uRDvw%40mail.gmail.com.

[Richard B. Winston]

Randy is a little premature in saying that OWHM2 is available in ModelMuse now. I submitted a version that supports OWHM2 for approval today. I hope to have it released soon.

To view this discussion on the web visit https://groups.google.com/d/msgid/modflow/96935D62-DACB-4EA6-9080-E7C529408319%40gmail.com.

[miyuru gunathilake]

Hi Randall, 

Is it possible for you to send a sample format of the rainfall files you used. 

Thank you in advance

Miyuru

To view this discussion on the web visit https://groups.google.com/d/msgid/modflow/96935D62-DACB-4EA6-9080-E7C529408319%40gmail.com.

[ashutosh singh]

Hi,

Just an honest request to have a couple of videos for advanced learners on how to incorporate surface water and unsaturated zone process in modflow-owhm using Modelmuse.

As far as I tested the executable it runs all of my modflow-nwt models so that is definitely a plus but it was taking longer time to run it. I was using serial runs.

Is it possible to just run the modflow-nwt model in parallel using owhm executable??

Regards

Ashutosh Singh 

To view this discussion on the web visit https://groups.google.com/d/msgid/modflow/61d561c9-0d41-477b-81ba-ebe295a1760e%40mindspring.com.

[Randall Hanson]

Dear Ashutosh

MF-OWHM contains all of MF-NWT plus additional NWT enhancements not available in MF-NWT, so you just need to run MF-OWHM.

Many other MF features have also been upgraded in MF-OWHM too.

Cheers 

Randy Hanson

One-Water Hydrologic

Sent from my iPhone

On Sep 15, 2023, at 12:24 AM, ashutosh singh <ashutoshs...@gmail.com> wrote:



To view this discussion on the web visit https://groups.google.com/d/msgid/modflow/CA%2BOP%2B8vYGMFey61sCce-XE6Y931J89Wq5qYEwpH0AhMpoUiUcA%40mail.gmail.com.

[Boyce, Scott E]

Here are some info links for MODFLOW-OHWM v2.3

Exe download and code:

https://code.usgs.gov/modflow/mf-owhm

Main Landing Website

https://www.usgs.gov/software/modflow-one-water-hydrologic-flow-model-mf-owhm 

Main Documentation (I usually recommend the first 20 pages and Appendix 4 for new users):

https://pubs.er.usgs.gov/publication/tm6A60 

I put together a FMP template, which is the package that handles precipitation arrays.

The header of the template talks about input structure and then follows the input bocks/descriptions (pretty much a summary of Appendix 6 of the tm6a60 report).

There are two versions of the header, one that is a regular text file and another that is html so you can see it with syntax highlighting in a web browser.

https://code.usgs.gov/modflow/mf-owhm/-/tree/main/doc/FMP_Template

For the sample files, check out the example directory at

https://code.usgs.gov/modflow/mf-owhm/-/tree/main/examples/mf-owhm

It only includes a simple use of precipitation but covers all the major features on setting up an OWHM model.

However, you may want to wait until ModelMuse support of OHWMv2 is approved for release.

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

For running the NWT solver in MODFLOW-OWHM, you have to disable the additional convergence metrics in OWHM to get the same runtime speed. OHWM added something called the Relative Volume Error that checks against each cells residual error divided by the volume of that cell. If any cell violates this limit, convergence is disabled even though RCLOSE and HCLOSE are satisfied. This was added to improve mass balances in the final solution. The two metrics that MODFLOW uses for convergence HCLOSE and RCLOSE end up in a delicate balance between being too small resulting in some time steps failing to converge or two large resulting in mass errors. We found in practice that the Relative Volume Error ends up improving the likelihood of time steps converging and lowering the mass balance. It's not a replacement for HCLOSE/RCLOSE but is just another requirement for a time step to be considered converged.

Regrettably the way MODFLOW is designed makes it difficult to calculate the mass balances during each solver iteration and making that a convergence criterion. We tried to add that as a option, but found it was going to require substantial refactoring of the base code, so it was abandoned in favor of relative volume errors.

I talk about Relative Volume Error and how to disable it in this forum post:

https://groups.google.com/g/modflow/c/ketiQRYu-f4/m/y8MtbiUsAgAJ   

There are some other minor differences in how OWHM handles the solvers under the hood that have a minor performance hit but provide the user with more information. For example, it is not compiled with as aggressive of compiler optimizations in order to provide Fortran info if there is a fatal error. There also a lot of additional warnings and catches that the code itself try's to intercept and help the user. Another biggie is that it calculates the mass errors for every time step, irrelevant of the Output Control (OC). Traditional MF only calculates the mass error if either the OC requests (I think its PRINT LIST) or if a time step fails to converge. OWHM checks it's for every time step and if it exceeds 5% raises a warning (the threshold can be changed with a keyword, check out all the BAS options at: https://code.usgs.gov/modflow/mf-owhm/-/blob/main/doc/Option_Block_Cheatsheets/BAS_Options_Recommended.bas). We added that feature when we realized that often time steps would converge, but had a terrible mass error and the user never knew unless they requested it in the OC.

One thing to point out, UPW (required by NWT) does give a slightly different solution compared to other solvers and LPF. I talk about that in the following post. The difference becomes exacerbated for large Specific Storage values as was demonstrated in the example here:

https://groups.google.com/g/modflow/c/29BH7rgNnd0/m/HVpNuihfGQAJ 

For the parallel question, not sure what you mean by that. Solving groundwater is inherently serial in nature. How people solve models in parallel is running multiple, independent models as in done in PEST.

We played a bit with parallel for reading the stress period input and solving the budgets at the end of the time step, but unfortuntely (at the time we tryed it) resulted in negligable improvements compared to the loss of an extra thread/cpu for running PEST on.

One thing that does help a lot with speed is buffering the in and output files. All files open in OWHM support the post-keyword BUFFER followed by the buffer size in kilobytes.

For example in the name file you can have something like:

LIST  55  list.txt   BUFFER 1024

which will write the list file to a 1 megabyte buffer and when its full write it to a file. This has the effect of writing large files like the LIST in 1MB chunks. For our models, we found we got about a 5% speed bump by doing this for large files like the Cell by Cell and LIST. The downside is that if there is a loss of power to the computer, the buffer is lost (cuz it never gets written out). We did find a buffer greater than 1024 does not improve speed and only waists ram. All files by default buffer at 32kb.

For input files, they are pre-read at the buffer size.

For example,

GHB  56  ghb.txt  BUFFER 64

if ghb.txt is 50kb in size, then the entire file is loaded into RAM at the start and the buffer is read instead of the actual file.

Hope that helps all out,

Scott

---

From: mod...@googlegroups.com <mod...@googlegroups.com> on behalf of ashutosh singh <ashutoshs...@gmail.com>
Sent: Thursday, September 14, 2023 11:24 PM
To: mod...@googlegroups.com <mod...@googlegroups.com>
Subject: [EXTERNAL] Re: [MODFLOW] Importing daily climate files into MODFLOW

 

 

This email has been received from outside of DOI - Use caution before clicking on links, opening attachments, or responding.

To view this discussion on the web visit https://groups.google.com/d/msgid/modflow/CA%2BOP%2B8vYGMFey61sCce-XE6Y931J89Wq5qYEwpH0AhMpoUiUcA%40mail.gmail.com.