Field data optimization

A discussion forum for Hydrus-1D users.
Post Reply
talat
Posts: 4
Joined: Tue Apr 07, 2020 2:21 pm
Location: Germany

Field data optimization

Post by talat » Sat May 23, 2020 11:06 pm

Dear Jirka,

I want to simulate water flow in the beech forest. The soil profile has five horizons. We measured water content and pressure head on each horizon every 15 min for two years. I read Hydrus1D user’s manual, hydrus1d tutorials, and some of your articles in this case but unfortunately, I can’t find the answers to my questions.

Could you please help me with these issues?

1- According to the number of rainy days per month, I divided data into dry and wet periods and then select first wet and dry periods separately for model calibration. But I wasn’t sure that this procedure is correct. How can I subset the data for calibration and validation of the model?

2- When did I optimize all of the soil hydraulic parameters for 5 depths I get an error that NpaD is exceed. How can I fix it?

3- In another way, I’ve selected only 3 parameters (α, n, and Ɵs) for optimization but the numerical solution isn’t converged.

Many thanks in advance.

Jirka
Posts: 4922
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Field data optimization

Post by Jirka » Sun May 24, 2020 7:39 pm

1. It is a common practice to use one season for calibration and one for validation.
2. For reason of uniqueness, we do not recommend to optimize many parameters simultaneously. See, for example, Hopmans, J. W., J. Šimůnek, N. Romano, and W. Durner, Inverse Modeling of Transient Water Flow, In: Methods of Soil Analysis, Part 4, Physical Methods, Chapter 3.6.2, Eds. J. H. Dane and G. C. Topp, Third edition, SSSA, Madison, WI, 963-1008, 2002.
3. Make sure that your optimized parameter values are meaningful. If needed, constrain the interval in which they can be optimized.
J.

talat
Posts: 4
Joined: Tue Apr 07, 2020 2:21 pm
Location: Germany

Re: Field data optimization

Post by talat » Mon Jun 01, 2020 6:33 pm

Dear Jirka,

Many thanks for your prompt answer and sorry for my delayed response. I read some parts of your suggested reference and did my project optimization. I used Rosetta for predicting the initial amount of soil hydraulic parameters and chose 2 or 3 parameters of α, n, Ks, Ɵs, and Ɵr and constrained them in meaningful amounts but the results aren’t satisfying. I attached my file to you. Could you please kindly check it and give me some advice?
Extra information regarding the project: I used throughfall and neglected evaporation because the soil surface is covered by leaf and plant residual.

All the best,
Talat
Attachments
optimization.rar
water flow in the beech forest
(358.61 KiB) Downloaded 2 times

Post Reply