Just a small nodal spacing is desirable, one would like to use small time steps to obtain an accurate solution as well. A good order-of-magnitude is to estimate the critical time step \(\Delta t_c\) with a formulae given by de Marsily (1986). For a homogeneous and isotropic aquifer \(\Delta t_c\) can be estimated by:

\( \seteqsection {12} \)

\( \seteqnumber {11} \)

\begin{align*} \Delta t_c=S \frac {\left ( \Delta x \times \Delta y \right )^2}{4T} \end{align*}

whereby \(\Delta x\) and \(\Delta y\) are the cell sizes of the model (m), \(S\) (-) is the porosity values and \(T\) is the transmissivity value (m\(^2\)/day). The result of the equation is given in the following figure: