Two algorithms have been tested against eachother, and against two external...

Two algorithms have been tested against eachother, and against two external inversions now. The result is that they give the same result as long as no localization is used. They both correspond completely now to a full kalman filter solution as well. They do however *not* correspond to the results from CT2009 because the fortran algorithm contained an error in the update loop of the simulated mean and deviations.