## Groudwater Level and Deep Drainage

A discussion forum for Hydrus-1D users.
julie j
Posts: 4
Joined: Mon Jan 23, 2006 3:44 pm
Location: France
Contact:

### Groudwater Level and Deep Drainage

Hello,

I have a problem building a model that includes an interface between the unsaturated and the saturated zones (from the surface down to the groundwater). Indeed, atmospheric data and observed water levels provide the required boundary conditions for the numerical model. As it is described in Exemple 2 of the tutorial, we must use the “deep drainage” condition.

Then we must specify the value of the parameters Aqh and Bqh, and the QWL0L, of the equation: q(h)=-Aqh.EXP.(Bqh.(h-GWL0L)) (as the exponential discharge of a watershed)

In the example 2, Aqh is negative, so q(h) must be positive…! And the GroundWaterLevel (GWL0L) is positive (+230cm); if this point is the bottom (lowest point) of the profile, and if the depth axis is positive upwards, shouldn’t GWL0L be negative ?

I then wonder how to determine the value of these parameters: Aqh and Bqh. Indeed, this requires to calculate q(h), at the bottom for instance, and then to evaluate these parameters. However I tried to figure q(h) at the bottom, for the exemple 2, from the values of h at different time steps, (data from the file Nod_inf.out), but I did not succeed to find the same results for the flux.
So, then, if I am not able to calculate the q(h), how can I evaluate Ahq and Bqh? Or how can I figure q(h) at the bottom of the profile?

At last, I thought that we could use the lower boundary condition “Variable Pressure head”; but, a model with the upper atmospheric BC and the initial condition which fixes the free surface above the bottom of the profile, although the profile is initially saturated above the bottom, the soil becomes dry way too fast.

I hope my questions are clear and that my English is a correct “Shakespeare English”.

Thank you very much for your help.

Julie

Jirka
Posts: 5495
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA
Julie,

The correct function is without minus in front of Aqh:
q(h)=Aqh.EXP.(Bqh.(h-GWL0L))

You should be able to draw this function using T_Level.out file using vBot and hBot.

Jirka

You can find more on this function in:

Hopmans, J. W. 1988. Treatment of spatially variable groundwater levels in one-dimensional stochastic unsaturated water-flow modeling. Agricultural Water Management 15:19-36.

It is shown in this paper how the coefficients are determined.
In concept, they can only be determined from measurements of groundwater table and drainage discharge of the field. If you have these data, you have a very nicely defined lower boundary condition that provides fluxes that are calibrated from measurements. Likely, the magntitudes of these parameters characterize the drainability (magnitude as a function of time) of the field, and is likely field-specific, depending on your parameters (e.g. drain spacing, soil type, macropores presence, drain material, lateral flow of groundwater, leakage to deeper groundwater, etc.).

julie j
Posts: 4
Joined: Mon Jan 23, 2006 3:44 pm
Location: France
Contact:
Jirka,

I have read the article, but I still have some questions..

I indeed can calculate the bottom flux, by the expenential law, with the hBot data from the output file.
However I want to calculate it from the Darcy law (as the bottom point belongs to the saturated zone, I use Ks and the pressure's gradient between the bottom point and another point which also belongs to the saturated zone). Using this law I can't find the vBot value. Why..??

Moreover what do you think about the "variable pressure head" lower BC instead of "deep drainage" lower BC?

Thank you very much...!!

Julie

Jirka
Posts: 5495
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA
Julie,

If you know the bottom flux, then obviously you can specify it as a constant or time-variable flux BC. The problem in most applications is that the flux at the bottom of the soil profile is not known, and that is why we offer all these other BCs, such as free drainage, deep drainage, seepage face, etc.

Jirka

rannu.bari@gmail.com
Posts: 12
Joined: Tue Jun 29, 2021 4:03 am
Location: Australia

### Re: Groudwater Level and Deep Drainage

Hi Jirka,
I have almost similar queries with this topic and that's why posting it here.

1) I am calculating recharge from the pond for 20 years where I considered 45 cm standing water during 10 months/year, soil depth 300 cm with 2 types. The groundwater level varies from 50 cm to 300 cm for the study area. Should I use the lower boundary condition as deep drainage or free drainage (I assumed free drainage but it's applicable if the water level is far from the domain)?

2) If the groundwater level from the T level output is hBot minus the height of the domain then
would it be like that (my apologies for the silly query) according to my information?

hBot=-150 cm/day, height of the domain (that is soil depth)= 300 cm,
so GWL= -150-300= -450 cm; I am confused that the soil depth will be minus or plus here

Thanks a lot for this platform and your continuous support.

Kind Regards,
Rahena

Jirka
Posts: 5495
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

### Re: Groudwater Level and Deep Drainage

Hi Rahena,

1. If you know the position of the groundwater, you should use the time variable-head BC at the bottom of the profile. If you have version 5 of HYDRUS, you can also use the option to interpolate linearly between specified values.

You should not use the Free Drainage BC at the bottom when the GWT is in the profile. The Free Drainage BC is only for conditions, when the GWT is deep below the soil profile bottom.

You can use the Deep Drainage BC (which assigns bottom flux as a function of the position of the GWT, which should be inside of the soil profile), but then you need to come up with the parameter for this equation.

2. hBot in the T_Level.out file is the position of the GWT above the bottom of the soil profile. Thus, indeed, the depth of the GWT is the depth of the soil profile minus this value. If the depth of the soil profile is 200 cm, and hBot is 50 cm, then the GWT is 150 cm below the soil surface.

J.

rannu.bari@gmail.com
Posts: 12
Joined: Tue Jun 29, 2021 4:03 am
Location: Australia

### Re: Groudwater Level and Deep Drainage

Hi Jirka,
Thank you.
I have weekly GWL and just for testing, I used those data with variable pressure head boundary condition keeping other conditions the same for rice crop and pond separately. But in both cases, the model shows the convergence error after several days of simulation.

Could you please give some tips (earlier I played with finer node, pressure head) to resolve that? In that case, if you need some specific information, I can provide that.

Kind Regards,
Rahena

Jirka
Posts: 5495
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

### Re: Groudwater Level and Deep Drainage

Rassam, D., J. Šimůnek, D. Mallants, and M. Th. van Genuchten, The HYDRUS-1D Software Package for Simulating the Movement of Water, Heat, and Multiple Solutes in Variably Saturated Media: Tutorial, Version 1.00, CSIRO Land and Water, Adelaide, Australia, 183 pp., ISBN 978-1-4863-1001-2, 2018. Appendices I, II, and III.

https://www.pc-progress.com/en/Default. ... torialBook

rannu.bari@gmail.com
Posts: 12
Joined: Tue Jun 29, 2021 4:03 am
Location: Australia

### Re: Groudwater Level and Deep Drainage

Hi Jirka,
Thank you.
I know I have some conceptual problems in understanding this convergence error. I am trying in different ways but can't figure out what reasons might be behind that.
Your tutorial book represents some basic steps to building a Hydrus model (including example 4). But I am not sure how these can enhance my understanding of convergence error.

Kind Regards,
Rahena

Jirka
Posts: 5495
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

### Re: Groudwater Level and Deep Drainage

The Richards equation is a highly nonlinear equation, which is solved in HYDRUS using the Picard iteration approach (which not always converges). You can read about this method in the HYDRUS manual. It may also help you to read the rules about spatial and temporal discretization (FAQ :
http://www.pc-progress.com/Documents/No ... zation.pdf
which are important for the convergence of the solution. J.

rannu.bari@gmail.com
Posts: 12
Joined: Tue Jun 29, 2021 4:03 am
Location: Australia

### Re: Groudwater Level and Deep Drainage

Hi Jirka,
Thank you for your continuous support.

I considered the lower boundary condition (variable pressure head with GWL, GWL varies from 150 cm-350 cm) for a paddy field with 50 mm ponding water for 125 days. The height of the model domain was 300 cm. There was irrigation event every 10 days (almost, depending on the water need) with 50 mm and on the first day, it was 122 mm for puddling condition.

I got the bottom flux that indicates the positive value above the surface (1st screenshot). Is this recharge also?
This pattern of recharge also does not follow with GWL pattern (2nd screenshot). So, assuming something wrong with that.
I attached herewith screenshots. Could you please have a look at it and help me to develop an understanding.

How can I get an example or manual for calibration?

Kind Regards,
Rahena
Attachments
image (1).png (32.53 KiB) Viewed 249 times
Screenshot 2022-06-22 123315.png (35.21 KiB) Viewed 249 times

Jirka
Posts: 5495
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

### Re: Groudwater Level and Deep Drainage

I’m attending a conference in Europe and thus will not have time to look at your problem.
Note that unless you use Version 5 of Hydrus, the time-variable BC change in steps (are not interpolated in time). As a result of a sudden change of the bottom BC (position of GWT), you can get abrupt changes in bottom fluxes. J.

rannu.bari@gmail.com
Posts: 12
Joined: Tue Jun 29, 2021 4:03 am
Location: Australia

### Re: Groudwater Level and Deep Drainage

Hi Jirka,
Thank you.
No worries. I can wait until your free time for a detailed explanation of my problem.

Is Hydrus-1D, version 5 free to use? It seems like I have to pay for that.

Kind Regards,
Rahena

Jirka
Posts: 5495
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

### Re: Groudwater Level and Deep Drainage

Version 5 is a commercial version. We will keep Version 4.17 of HYDRUS-1D in the public domain as it is, but any new developments will be implemented only in Version 5 (which will be commercial). J.

rannu.bari@gmail.com
Posts: 12
Joined: Tue Jun 29, 2021 4:03 am
Location: Australia

### Re: Groudwater Level and Deep Drainage

Hi Jirka,
I used interpolated GWL for that simulation.
I attached herewith my sample Hydrus file.
Could you please have a look at it once you are available to help me to find out the positive recharge issue that I mentioned in the previous comment (with screenshots)?

Kind Regards,
Rahena
Attachments
sample hydrus file.zip