To overcome the inconsistencies of the historic IGS products due to modeling and reference frame changes the IGS initiated a reprocessing project. The goal of this project is the generation of a consistent and homogeneous set of IGS products starting with 1 January 1994. This contribution will discuss the type and temporal resolution of the reprocessed products. To generate a consistent set of reprocessed IGS products a number of common standards should be fulfilled by the analysis centers participating in the IGS reprocessing effort. Some of these issues have already been discussed and a consensus has been reached which will be summarized in this contribution.

GPS/GLONASS P1-P2 as well as GPS P1-C1 differential code bias (DCB) values are computed at the CODE analysis center on a regular basis. The two methods that are used at CODE for the retrieval of P1-C1 values are explained in detail. The second method even leads to an "ambiguity-fixed" DCB product. We present various time series of DCB results. It is intended to address problems related to the P2-C2 issue.

The IGS reprocessing campaign will benefit from the use of homogeneous processing strategies and models by the participating Analysis Centers for the history of GPS observing since 1994. This will provide a greatly improved and more useful product set for users. As part of the updating process being made in anticipation of this effort, a review of recent and planned changes for the IERS Conventions will be presented. Emphasis will be placed on those model changes for which a common IGS approach is most desirable.

A reprocessing of a global GPS network has been carried out at the Universities of Technology in Dresden and Munich. Up to now almost 12 years of data have been processed in a joint effort. The major benefit from a reprocessing of all observation data can be seen in the opportunity to apply the latest models and processing strategies in the data analysis in order to obtain homogeneous and consistent parameter time series.

In this context we introduce the implemented data processing scheme. The incorporated strategy and model refinements (e.g., the isobaric mapping function based on weather model data and the application of 2nd- and 3rd-order ionospheric corrections) as well as their benefits are pointed out. In addition we present results for different parameter types with a focus on coordinate time series in order to demonstrate the homogeneity of the reprocessed solution.

In the GPS data analysis, due to computing time and limited computer memory, large networks are usually divided into sub-networks, processed separately and then combined on the normal equation (NEQ) level. For the same reason, an integrated analysis of ground- and space-based GPS data can only be carried out on computers of high performance. This conflict between high-quality products and computer resources becomes more and more critical with the growing ground networks, the increasing number of low earth orbiting (LEO) satellites and the upcoming GALILEO system. One of the major reasons for this conflict is that a huge number of ambiguity parameters have to be kept in the NEQ for integer ambiguity fixing.

In the presented strategy, parameters defined over a time interval shorter than the whole session length, for example, ambiguity and zenith tropospheric delay parameters, can be eliminated immediately after all related observations have been contributed to the NEQ even for ambiguity-fixed solution. It reduces the requested memory and computing time dramatically from that of the current methods.

We will demonstrate that with the new strategy a huge IGS network up to 250 stations can be analyzed on a PC with 1 GB memory, and that the computing time can be reduced to one third of the current method. We will also show the differences between our IGS products from the integrated estimation and the sub-network strategy.

The purpose of this talk is to summarize the present status and expected results in the reprocessing of the IGS Global Ionospheric Vertical Total Electron Content Maps (GIMs), as a result of the official IGS call, happened few months ago.

Indeed, the IGS GIMs are being computed since June 1998, as the weighted combination of the different GIMs processed by the involved agencies (presently CODE, ESA, JPL and UPC). During this time, the ionospheric techniques used by the agencies have experienced significant improvements: tomography, physical models, increase of temporal resolution, better interpolation techniques, are some relevant improvements, introduced by different analysis centers. In such situation the backward recomputation of IGS GIMs, proposed in the context of the general IGS campaign, can provide a significantly better ionospheric product from IGS.

To illustrate that, we will summarized as well an approach based on the UPC interpolation algorithm, adapted from the Kriging technique, which allows not only the improvement of its own estimation, but the improvement of the ionospheric maps computed by other agencies as well, in a very simple and straightforward way. This fact could be quite convenient for the IGS reprocessing task. The first results obtained with this technique in the computation of the GIMs for the year 2000, show an improvement of 10% or more for practically all the analysis centers.

This presentation will summarize the processing strategies and models of the analysis centers who have already agreed to participate in the IGS reprocessing effort. Guidelines for the station selection of the individual analysis centers will be given to achieve an optimal coverage of the IGS network by the reprocessing. The options for the standards which have not been agreed untill now will be discussed and finalized.