First-order continuation method for steady-state variably saturated groundwater flow modeling

Denis Anuprienko
DOI: https://doi.org/10.48550/arXiv.2201.05202
2022-01-14
Abstract:Recently, the nonlinearity continuation method has been used to numerically solve boundary value problems for steady-state Richards equation. The method can be considered as a predictor-corrector procedure with the simplest form which has been applied to date having a trivial, zeroth-order predictor. In this article, effect of a more sophisticated predictor technique is examined. Numerical experiments are performed with finite volume and mimetic finite difference discretizations on various problems, including real-life examples.
Numerical Analysis
What problem does this paper attempt to address?
The main problem that this paper attempts to solve is to improve the nonlinearity continuation method for solving the boundary - value problems of the steady - state Richards equation. Specifically, the author improves the performance of this method by introducing a more complex predictor step, namely the first - order predictor. In addition, the author also extends the discretization framework from being limited to the finite volume schemes to including the mimetic finite difference method in order to test the performance of the nonlinear continuation procedure on different types of nonlinear systems. ### Main research contents: 1. **Application of the first - order predictor**: The paper proposes and evaluates a more complex first - order predictor, which can significantly reduce the continuation steps and the number of Newton iterations, thus speeding up the calculation. 2. **Extension of the discretization method**: In addition to the traditional finite volume method, the mimetic finite difference method is also introduced to verify the applicability and efficiency of the nonlinear continuation method in different types of discretization schemes. ### Key formulas: - **Richards equation**: \[ -\nabla \cdot (K_r(h) K \nabla h) = Q \] where: - \( h \) is the hydraulic head. - \( K_r(h) \) is the relative permeability. - \( K \) is the hydraulic conductivity tensor. - \( Q \) is the source and sink terms. - **Van Genuchten - Mualem constitutive relations**: \[ \theta(\psi) = \theta_r + (\theta_s - \theta_r) \left(1 + |\alpha \psi|^n\right)^{-m} \] \[ K_r(\theta) = S_e^{1/2} \left(1 - \left(1 - S_e^{1/m}\right)^m\right)^2 \] where: - \( \theta_s \) is the water content at saturation. - \( \theta_r \) is the residual water content. - \( S_e=\frac{\theta - \theta_r}{\theta_s - \theta_r} \) is the effective saturation. - \( n > 1 \), \( m = 1-\frac{1}{n} \), \( \alpha \) is a model parameter. Through these improvements, the paper aims to improve the efficiency and stability of the nonlinear continuation method in solving variably - saturated groundwater flow problems.