Page 2 of 2

Re: HYDRUS 1D Model Fails to Converge

Posted: Fri Mar 27, 2020 9:08 am
by hhfang
Dear professor Jirka,
Thank you very much for the reply! Actually I tried, but I got two different output in the end (as shown in the Fig.1 and Fig.2).
I ran HYDRUS 1D with a single layer as the reference case. I added Alfa2 , n2, but set w2 =0 in the SELECTOR.IN file, and ran HYDRUS again with all exact the same input files and settings (Fig.3). I compared the results and found that they are different in the end. But I also notice that the observed node in the deep layer(131 and 211) already show difference in the beginning at the time 10 sec (Fig.1). Do I miss or misunderstood something?
Thanks a lot for the help!

Fig.1
out1.PNG
out1.PNG (361.49 KiB) Viewed 598 times
Fig2.
out2.PNG
out2.PNG (102.16 KiB) Viewed 598 times
Fig3.
selectorin.PNG
selectorin.PNG (100.71 KiB) Viewed 598 times

Re: HYDRUS 1D Model Fails to Converge

Posted: Fri Mar 27, 2020 2:53 pm
by Jirka
From your input I can see that you use either the Kosugi (Model=3) or Durner dual-porosity (Model=5) models. Durner model is obviously not based on the Kosugi's model, but on the van Genuchten-Mualem model. Thus, if you want to compare these two simulations, you need to use consistent models.

J.

Re: HYDRUS 1D Model Fails to Converge

Posted: Fri Mar 27, 2020 5:20 pm
by hhfang
Thank you very much!