Nonlinearity continuation method for steady-state groundwater flow modeling in variably saturated conditions

Denis Anuprienko,Ivan Kapyrin

DOI: https://doi.org/10.48550/arXiv.2008.00730

2020-08-03

Abstract:Application of nonlinearity continuation method to numerical solution of steady-state groundwater flow in variably saturated conditions is presented. In order to solve the system of nonlinear equations obtained by finite volume discretization of steady-state Richards equation, a series of problems with increasing nonlinearity are solved using the Newton method. This approach is compared to pseudo-transient method on several test cases, including real site problems and involving parallel computations.

Numerical Analysis

What problem does this paper attempt to address?

This paper attempts to solve the nonlinear problems encountered in the numerical simulation of steady - state groundwater flow under variable - saturation conditions. Specifically, the author focuses on how to effectively solve the nonlinear equation systems obtained by finite - volume discretization, which are derived from the Richards equation describing steady - state groundwater flow.

### Main problems
1. **Selection of initial guess values**: For complex heterogeneous and anisotropic media, especially in dry soil conditions or with highly nonlinear constitutive relationships, the Newton method requires an initial guess value close enough to the real solution to converge. However, in practical applications, it is not easy to find such an initial guess value.
2. **Limitations of the pseudo - transient method**: Although the pseudo - transient method can gradually bring the system to a steady state by introducing the time dimension, this method may encounter severe time - step limitations when dealing with highly heterogeneous and relatively dry soils, and has a high computational complexity.

### Proposed solutions
To solve the above problems, the author proposes a **Nonlinearity Continuation Method**. This method obtains initial guess values suitable for the Newton method by gradually increasing the nonlinearity of the problem, thereby avoiding the convergence difficulties that may be encountered when directly solving highly nonlinear problems.

### Method principle
- Introduce a continuation parameter \( q \) to control the nonlinearity of the problem:
\[
-\nabla \cdot (K(h, q) K \nabla h) = Q
\]
where \( K(h, q) \) is a function that satisfies the following conditions:
\[
K(h, 1) \equiv K_r(h), \quad K(h, 0) \equiv 1
\]
Thus, when \( q = 1 \), the equation becomes the original nonlinear Richards equation; when \( q = 0 \), the equation becomes a simple linear equation.
- Start from \( q = 0 \), gradually increase the value of \( q \) until \( q = 1 \), and use the previous solution as the initial guess value each time to solve the current nonlinear equation by the Newton method.

### Experimental results
Through tests on several model problems and actual site problems, the experimental results show that the Nonlinearity Continuation Method has fewer Newton iteration times and faster calculation speed compared with the pseudo - transient method, especially when using the more complex MPFA discretization scheme.

In conclusion, this paper aims to provide an effective solution through the Nonlinearity Continuation Method to overcome the nonlinear solution problems in the numerical simulation of steady - state groundwater flow under variable - saturation conditions.