Lange's Covariance Matrix (LCM):
which is the inverse of:
and that renders the exact precision
of Wolf's analytic solution:
to Helmert's problem:
which canonical form can always be achieved, for example,
through performing a generalized Canonical Correlation Analysis (gCCA)
on the observation errors.
It would though be advantageous to find a most meaningful way of clustering the data into mutually uncorrelated blocks
which would then make the best sense also physically.
This semi-analytic inversion is known in Geodesy as the Helmert-Wolf blocking
(HWB) method and
to Satellite Geodesy.
The LCM formula above is seminal to the Fast
Filtering (FKF) method which, in turn, is a "must"
to all mobile positioning, navigation, forecasting and control appliances as well as
to Numerical Weather Prediction
systems that comply with the highest precision, integrity and safety rules and standards.
* Last revised: August 19, 2005