Allow for a more pessimistic model of test-mass noise

The current model for test-mass noise nominally extends down to the low limit of the LISA band, i.e., 1E-4 Hz.

I need a model to go to lower frequencies (a few 1E-5 Hz). I can use the current model, but to ensure I am not unreasonably optimistic, I would like to add low-frequency relaxation to the current test-mass noise model.