Fix TRM restart

Fixes a bug introduced in !4316 (merged).

TRM restart did not work since !4316 (merged)! However, OGS should have aborted if somebody had tried to restart a TRM simulation after !4316 (merged).

1. Feature description was added to the changelog
2. Tests covering your feature were added?
3. Any new feature or behavior change was documented?

I added restart variants of the TRM and RM Liakopoulos tests to tests the restart. The vertical profiles through the column are as follows (TRM case):

I.e., optically, there is a perfect match between the reference result and the restarted simulation.

However, the error tolerances had to be increased quite a bit for the restart case, namely from

    liakopoulos_t_7200.vtu liakopoulos_t_7200.vtu sigma sigma 1e-9 1e-12
    liakopoulos_t_7200.vtu liakopoulos_t_7200.vtu displacement displacement 1e-10 1e-12
    liakopoulos_t_7200.vtu liakopoulos_t_7200.vtu saturation saturation 1e-10 1e-12

to

    liakopoulos_t_7200.vtu liakopoulos_restart_t_7200.vtu sigma sigma 1.5 0
    liakopoulos_t_7200.vtu liakopoulos_restart_t_7200.vtu displacement displacement 3.4e-7 0
    liakopoulos_t_7200.vtu liakopoulos_restart_t_7200.vtu saturation saturation 4e-5 0

which is a bit of a surprise to me. I assumed that restart should produce identical results compared to the case without restart. But apparently there are some small differences.

The situation is the same for RM and TRM. The tolerances had to be increased to the same values for both processes.