Sect_9.3.2

9.3.2 Multi-Session Network Processing

OBTAINING A MULTI-SESSION SOLUTION

As in the case of single session solution, there are a several strategies for obtaining a multi-session solution:

Using multi-session GPS data reduction software. This ensures that stations that are common to more than one session are correctly weighted in the solution, and that datum transfer is carried out within the matrix operations of the Least Squares adjustment. High precision scientific GPS software invariably has a multi-session capability.

Secondary network adjustment software uses the results of individual session solutions as input. This is the same software that is used at the single session processing level. General geodetic adjustment packages can, in addition to processing GPS baselines as observations, permit input of conventional survey observations such as distances, theodolite directions, etc. There are many other software packages that will handle only the GPS baselines as input.

In the case of single session adjustments, it is clear that a secondary adjustment of single baselines is sub-optimal compared with a multi-baseline GPS reduction. This is not so obvious in the case for multi-session adjustments because of the uncorrelated nature of double-differences between sessions. However, there are circumstances where the functional correlation of parameters between sessions would suggest that a simultaneous reduction of the GPS phase data collected over a number of sessions is the superior approach. Two can be mentioned:

In the case of the adjustment of common parameters, such as satellite orbital elements, that span several sessions. Commercial GPS software packages do not have this capability. Packages designed for GPS geodesy applications invariably have a multi-session processing capability because of the orbit adjustment requirement.
Where many redundancies exist, by virtue of more than one common station linking sessions and/or the reoccupation of several stations in different sessions (see Figure below), a secondary network adjustment is sub-optimal.

A multi-session station configuration with redundancies.

However, it must be emphasised, perhaps even more than in the case of single session processing, that in practice there is often little discernible effect, at the few parts per million accuracy level, on GPS solutions, between: (a) processing single sessions, then combining them in a secondary adjustment, and (b) processing all observations from several sessions in one multi-session GPS phase adjustment. In fact, there are advantages in combining both approaches where possible. For example, single session solutions could be obtained as an initial office procedure to verify the data quality, followed by a rigorous multi-session solution only when the analyst is satisfied that the network "fits" adequately.

Back to Chapter 9 Contents / Next Topic / Previous Topic