Bugfix: test saturation values for table order.

[atgeirr]

As discussed in #4081, fixing a bug potentially turning the saturation tables.

[atgeirr]

jenkins build this please

[atgeirr]

The high number of test failures is a bit disheartening, but on the other hand the interpolation seems to be robust with respect to the order of the saturation values, so it should not matter significantly (i.e. getting wrong results) if order is flipped or not.

[blattms]

What does the Sw stand for?

[atgeirr]

What does the Sw stand for?

I think in this particular context it should probably be just s, for saturation. While not 100% sure of the lineage of the code it probably stood for "saturation, wetting" in the original context. Which of course is very easy to confuse with water saturation.

[blattms]

So values is the input to the tabulated functions and Swvalues are used for interpolate the the result for a particular value.

[atgeirr]

So values is the input to the tabulated functions and Swvalues are used for interpolate the the result for a particular value.

This class gives a function f(x) by piecewise linear interpolation, where the (x_i, f(x_i)) data are (swValues[i], values[i]). Which is maybe what you said, but I was not sure...