Issue with the model convergence for water flow

A discussion forum for Hydrus-1D users.
Post Reply
Posts: 16
Joined: Thu Dec 17, 2020 10:11 pm
Location: USA

Issue with the model convergence for water flow

Post by Hydraus_1d_RD » Fri Jan 01, 2021 12:19 am

Dear all,
I’m presently trying to figure out how to address a convergence related issue for my soil moisture flow problem. I have a column with 4 soil layers (total thickness 61 cm) and one rock layer (total thickness 1 cm). I discretized the whole soil column into 0.2 cm intervals and I’m using the following soil layer properties:
Mat θr θs α (1/cm) n (-) Ks (cm/day) l
1 0.1 0.39 0.059 1.48 23.33 0.5
2 0.068 0.38 0.008 1.09 7.86 0.5
3 0.068 0.38 0.008 1.09 7.86 0.5
4 0.068 0.38 0.008 1.09 7.86 0.5
5 0.02 0.17 0.05 1.67 0.30 0.50

As layers 2,3, and 4 are clay layers, I’m using the following option for the single porosity models:
Van Genuchten -Mualem model: with air-entry value of -2 cm

At present, I’m not considering any hysteresis.

I’m using the following time information:
Total time =2190 days
Initial time step (days)= 1e-5
Minimum time step= 1e-5

As my objective is to run a transient model for three layers, I used the average conditions for 6 years (from 3 years of data) for precipitation and meteorological data from the precipitation and meteorological time series data for the three year period.
My problem is that my model does not converge for many internal time steps. For example, the run_inf.out file shows “F” for many time steps. The A-Level.out file shows surface pressure head values sometimes ********* or sometime -1e5. Finally, the Balance.out file shows sometimes very high relative mass balance errors (sometimes >100 %).

Would you please let me know if I may request any suggestions or tips for moving forward?

Thank you!
Happy New Year!

Post Reply