7.1.6 Introduction

DATA PROCESSING CONSIDERATIONS

Among the most important considerations for the reliable processing of GPS data therefore are:

Correct field procedures to ensure that good quality data is collected in the field (Chapter 10): for example, avoiding multipath inducing environments, correct measurement of antenna height, adequate battery power, no signal obstructions (trees, etc.) or interference (microwave transmissions, etc.), appropriate length observation sessions, etc.

Clean data that has been appropriately pre-processed in order to detect and repair any cycle slips.

Appropriate observation modelling and processing software: matched to the accuracy required so that the correct strategy is used to account for the GPS measurement biases, and in particular the cycle ambiguities.

Appropriate processing procedures: recognises that GPS phase data reduction involves a number of steps executed in sequence.

The latter two issues are discussed briefly below.

Degree of Sophistication of Data Modelling and Processing

The GPS relative positioning technique is employed for a wide range of applications. As a starting point, in order to sort out the variety of GPS modelling and software options that are available, the surveying applications may be categorised according to three general ranges of accuracy:

	Class A (Scientific):	better than 1 ppm
	Class B (Geodetic):	1 to 10 ppm
	Class C (General Surveying):	lower than 10 ppm.

Class A surveys are primarily undertaken for precise engineering, deformation analysis, and geodynamic applications. Class B surveys include control surveys undertaken for the purposes of geodetic network densification, mapping, and for resource development applications. Class C surveys primarily encompass those lower accuracy surveys undertaken for urban, cadastral, GIS and general-purpose survey applications. Users in the latter two categories comprise the majority of the GPS survey user community, and it is the applications in these categories that are collectively referred to in these notes as "GPS surveying". In section 7.2.2 the GPS observation modelling options appropriate for GPS surveying are summarised.

Steps in GPS Data Processing

For the purpose of discussion, the data pre-processing, initial analysis and final adjustment steps as they relate to a single session are introduced here. These are described in the remaining sections of Chapter 7, in Chapter 8, and in parts of Chapter 9. The steps relating to the combination of session adjustments and the integration of the results with terrestrial control are discussed separately in Chapter 9, Chapter 11 and Chapter 12.

Data Pre-Processing

Pre-processing encompasses a number of specific tasks:

• Initial data transfer and decoding.
• Data screening and editing.
• Data reporting and database creation and entry.
• Point positioning using pseudo-range data.

These tasks may be carried out on a single-station basis, and can therefore be carried out in the field office. The result should be a set of appropriately formatted and pre-processed (though not entirely "clean") observations, together with ephemeris information and approximate station coordinates.

Initial Data Analysis

These tasks are carried out as a prelude to the final phase data adjustment, as soon as data from two or more GPS receivers is brought together in the field office:

• Cycle slip detection and repair.
• Preliminary baseline solution based on triple-differenced phase data or double-differenced pseudo-range data (if of suitable quality).

The former is discussed in section 7.3, together with the pre-processing procedures. The latter, as it is based on the principles of baseline determination, is discussed in section 8.1, together with double-differenced phase solutions. The result is a "clean" set of observations, and very good apriori station coordinates.

Final Adjustment

The adjustment of the pre-processed GPS observations similarly encompasses a number of tasks:

• The formation of the phase data differences.
• Definition of the apriori weight matrices.
• Estimation of relative station coordinates.
• Estimation of carrier beat phase ambiguities and fixing them (if possible) to integers.
• Output of estimated parameters and aposteriori covariance matrix.

The actual adjustment of GPS observations can be performed in two ways:

• by combining single baseline solutions into a network adjustment of the baseline components, or
• by a direct, simultaneous solution involving all the GPS observations for a session, or complete network.

Various types of differenced (including non-differenced) observables can be employed in GPS data processing (section 7.2.2). Each of these observable types have certain advantages and disadvantages with respect to bias reduction, parameter sensitivity, error modelling, and processing efficiency. The capability to exploit the integer nature of the carrier phase ambiguity terms is also an important consideration in the final adjustment stage. Methods of integer ambiguity resolution appropriate to the different observable types (apart from triple-differences), and for single and dual-frequency operations are needed (Chapter 8).

Back To Chapter 7 Contents / Next Topic / Previous Topic

© Chris Rizos, SNAP-UNSW, 1999