Handling mass gain in calcBgGD

[sashahafner]

With b3cc327 I removed error and set mass change mass.tot to 0.
But that will bias results caused by noise in last digit.
So with caeb5d1 I left these values negative, but that throws an error from one of the called functions that expects positive values for mass argument.
I think I should change that in the check... call in whatever that function is.