08.06.2013 Views

Bernese GPS Software Version 5.0 - Bernese GNSS Software

Bernese GPS Software Version 5.0 - Bernese GNSS Software

Bernese GPS Software Version 5.0 - Bernese GNSS Software

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

9.3 Manipulation of Normal Equations<br />

ADDNEQ2 this manipulation is performed based on the information contained in the station<br />

information file (see Section 9.4.6). Changes are notified to the user in the program output<br />

file.<br />

The station information file allows a consistent change of station and equipment names as<br />

well as of antenna eccentricities which may be very useful, e.g., for reprocessing. A change<br />

in the antenna position is taken into account by transforming the a priori station position<br />

information (see Section 9.3.4). On the other hand, when changing the antenna name only<br />

the antenna name is changed. It is not possible to consider an alternative antenna phase<br />

center variation model on this level of analysis. Some auxiliary parameter information can,<br />

however, not be changed at normal equation level. Examples are tropospheric mapping<br />

function or satellite antenna group definition (see Section 9.3.11).<br />

9.3.2 Rescaling the Normal Equation Matrices<br />

Rescaling of normal equations is important, e.g., if two or more normal equation systems<br />

have to be combined, where each of them stems from a different processing software or<br />

strategy. A typical example is given by the use of different sampling rates for <strong>GNSS</strong> observations.<br />

Due to the time-correlations of <strong>GNSS</strong> observations, the results remain (almost) the<br />

same, but the variance-covariance matrices change if different sampling rates are used.<br />

The problem is solved by the following transformation: assuming the original normal equation<br />

system<br />

N p = b , (9.15)<br />

the new system reads as<br />

κ N p = κ b , (9.16)<br />

where κ is the a priori scale factor. Its statistical meaning is the ratio of the variances:<br />

κ = σ2 old<br />

σ 2 new<br />

. (9.17)<br />

In program ADDNEQ2 the rescaling is performed if the a priori information on the scale<br />

factor is provided in a WGT file (option “Variance rescaling factors” see Section 22.10.5).<br />

9.3.3 A Priori Transformation of Coordinates into a Different Reference Frame<br />

A parameter transformation must be used for the transition between two reference frames.<br />

Let us assume that we want to combine normal equations stemming from solutions with<br />

fixed coordinates in, e.g., ITRF97 (denoted by index “97”) with a second set of normal<br />

equations where the ITRF2000 (denoted further by index “00”) was used. Both reference<br />

frames are related by the 7-parameter transformation (rates neglected)<br />

⎛<br />

⎜<br />

⎝<br />

Xi<br />

Yi<br />

Zi<br />

⎞<br />

⎟<br />

⎠<br />

00<br />

= (1 + µ)<br />

⎛<br />

⎜<br />

⎝<br />

1 γ −β<br />

−γ 1 α<br />

β −α 1<br />

⎞<br />

⎟<br />

⎠<br />

<br />

R<br />

⎛<br />

⎜<br />

⎝<br />

Xi<br />

Yi<br />

Zi<br />

⎞<br />

⎟<br />

⎠<br />

97<br />

⎛ ⎞<br />

∆X<br />

⎜ ⎟<br />

+ ⎝ ∆Y ⎠<br />

<br />

∆Z<br />

<br />

∆<br />

<br />

. (9.18)<br />

<strong>Bernese</strong> <strong>GPS</strong> <strong>Software</strong> <strong>Version</strong> <strong>5.0</strong> Page 187

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!