Odd flux patterns in soil column model

A discussion forum for users of the new HYDRUS 2D/3D. HYDRUS is a software package for simulating water, heat and solute movement in two- and three-dimensional variably saturated media. Happy Posting!
Post Reply
jscherberg
Posts: 8
Joined: Fri Aug 18, 2017 4:31 pm
Location: USA

Odd flux patterns in soil column model

Post by jscherberg » Fri Aug 18, 2017 4:51 pm

I'm working on a 2D model with a 10m x 1cm domain with an atmospheric upper boundary and a free drainage lower boundary, no flow on the sides. There are 2 material types with a cover layer of alluvial material over other material that can be more or less conductive, depending on the model variation. I am running 50 years of hourly data for an area where there are infrequent, but high intensity precipitation events and there is sparse vegetation. We have done a spinup run to get initial conditions. Our interest is in how much of the precipitation percolates. We are finding that all of the precip is consumed by ET, and with no vegetation evaporation alone removes most of the water and looking at cumulative flux at 1 meter increments shows alternating areas where there is upward and downward flow with a small amount of outflow at the bottom boundary. How can there be net downward flux in one region if the region above it has net upward flux given that initial conditions are dry and there is not much water in storage? Thanks for any explanations
Jake

Jirka
Posts: 4712
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Odd flux patterns in soil column model

Post by Jirka » Fri Aug 18, 2017 8:30 pm

Do you have a 10 m deep profile? Are you actually running a 1D problem?

The free drainage BC can result only in the downward flow. If you want to have precise fluxes at different depths, you should use meshlines and calculate fluxes across these meshlines. Additionally, you should divide the domain into subregions (between meshlines) so that you can check the mass balance. The mass balance error in each of these subregions should not be larger than the one reported in Mass Balance Information.

J.

jscherberg
Posts: 8
Joined: Fri Aug 18, 2017 4:31 pm
Location: USA

Re: Odd flux patterns in soil column model

Post by jscherberg » Fri Aug 18, 2017 11:16 pm

Thanks Jirka for the quick reply,
this is essentially a 1D problem the way we are approaching it which is why the domain is 1 cm wide. I am using flux lines and subregions. It is the odd outputs for flux from the flux lines that is the basis for my question. The only water source is precip and whatever small amount of water is present from initial conditions. Based on results from the cumulative flux at the meshlines there is downward flow at some portions of the domain and upwards flow at others. I can't see how this can be the case; for example in one model the cumulative fluxes, in cm at various depths are as follows:

30.5 cm 4.622
61 cm -10.06
1 m 9.538
2 m 8.562
3 m 7.358
4 m 5.893
5 m -4.245
6 m 2.555
7 m 0.6988
8 m -1.221

the bottom boundary flux is 5.36 cm. I don't understand how the flux can be negative in some places and positive in others. In this scenario rooting only goes to 1 ft below surface.

thanks,
Jake

Jirka
Posts: 4712
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Odd flux patterns in soil column model

Post by Jirka » Sat Aug 19, 2017 12:25 am

Make sure that you are interpreting the fluxes correctly. Note that meshlines are oriented depending on how they are defined, i.e., which node is the first and which one is the last. There is no convention as in 1D that fluxes up are positive and down are negative (that does not make sense in multiple dimensions). Here, it all depends on the orientation of the meshline and in what sequence you define it. It seems to me that you simply define the meshline at depth of 5 m in the opposite direction. J.

jscherberg
Posts: 8
Joined: Fri Aug 18, 2017 4:31 pm
Location: USA

Re: Odd flux patterns in soil column model

Post by jscherberg » Sat Aug 19, 2017 1:29 am

Thanks Jirka. I did not realize that. I appreciate your responding so quickly!
Jake

Post Reply