干旱区生态环境知识资源中心

A precise and accurate position of a GNSS network can be obtained by using many software. Each software implements different approaches and allows the choice of many strategies of computation. Each strategy require different complexity levels of setting up and "time to process". The aim of the paper is to evaluate the differences of using different approaches and software in the adjustment of a CORS network.

In this note three different software have been employed, using both the "differenced" (Bernese V.5.0 and Gamit-Globk) and "undifferenced" (Gipsy-Oasis II) approach. The dataset derives from a network of more than 140 GPS stations observed during an interval of 212 days: the network covers the whole Italian territory.

The aim of this research is the evaluation of the differences in term of obtained results (position and velocity parameters estimation) and efficiency, by using the three software in different configurations starting from a common dataset of data and global parameters.

Starting from quite the same parameters for all software the obtained differences, in term of average solution estimation, are comparable at few millimeter level. In particular, the solutions obtained by differenced approaches are very closed; in some case however, PPP (Precise Point Positioning) solutions evidence a few millimeter bias in the East Component not present in the other components. By the way, the velocity parameters estimation is quite the same for all the obtained solutions.

Considering the "time to process" aspect, when it has been possible more parallel thread have been performed: This has allowed a better use of the CPU and the present processors in the computer. In particular the PPP method (Gipsy solution) permits a very high flexibility that allows very high performances.