CSRS-PPP version 3: PPP-AR

I have been talking about it for years now, but it is finally coming: in October 2020, CSRS-PPP will offer precise point positioning with ambiguity resolution (PPP-AR) solutions for data collected on or after 1 January 2018.


As announced yesterday on the NRCan website, CSRS-PPP version 3 is expected to be launched next month. The main feature of this version is the implementation of PPP-AR algorithms. The NRCan webpage contains a link to a tutorial that explains the main changes to be expected from version 3. This blog post highlights some key points of this document (and adds a few extra comments).


Enabling PPP-AR has proven to be a challenging project since many components had to be developed to support this new functionality. Satellite orbit and clock products now require consistent code and phase biases preserving the integer nature of the ambiguities at the user end. NRCan thus had to modernize its lines of products to generate ultra-rapid, rapid and final “integer clocks.” The final products still use the IGS combined orbits, but an in-house clock combination replaces the IGS clock solution; it leverages the integer clocks of NRCan (rapid), CODE and CNES/CLS (GRG) while also incorporating contributions from other analysis centers (ESA, GFZ, JPL) in float mode. This new combination process is described by Banville et al. (2020). A differential code bias combination is also performed based on the solutions of CAS, CODE, DLR and NRCan.


The datasets submitted to our service vary widely in terms of quality: from geodetic receivers to low-cost receivers; from clear skies to urban canyons; from multi-day sessions to a single epoch of data; etc. Still, all datasets are processed using exactly the same PPP-AR methodology. To minimize potential issues with ambiguity resolution, the following precautions were taken: ambiguity resolution is disabled for single-frequency receivers, cross-correlation receivers, GLONASS ambiguities or ambiguities observed for less than 5 minutes. Partial ambiguity resolution is performed and the validity of the stochastic model plays a role in deciding whether or not to resolve ambiguities. Overall, I believe the algorithms are rather robust, but there are always corner cases where incorrect ambiguity resolution could occur. A paper describing the client-side PPP-AR algorithms is in preparation and should be submitted to a journal as soon as CSRS-PPP version 3 is launched.


Some of the outputs have changed in version 3. The receiver clock and tropospheric zenith delay estimates have been moved from the ‘pos’ file to the ‘clk’ and ‘tro’ files, which use formats adopted by the IGS. The PDF report also includes an ambiguity status plot (see figure below). This plot is my favorite validation tool to assess the quality of a dataset. The green lines indicate fixed ambiguities; olive are float ambiguities, cyan are datum/reference ambiguities, and red bars indicate an ambiguity reset. The tutorial available on the NRCan website contains other examples where ambiguity resolution was not as successful.

Figure 1: ambiguity status plot from CSRS-PPP version 3 PDF report


CSRS-PPP version 3 is the second step in our modernization plan. The first step was a change in PPP software in August 2018. The third step will include precise slant ionospheric delay corrections for Canadian users and is expected in 2021. Adding support for GPS L5 and Galileo will come after, most likely in 2022-2023.



Banville S, Geng J, Loyer S, Schaer S, Springer S, Strasser S (2020) On the interoperability of IGS products for precise point positioning with ambiguity resolution. J Geod 94. doi:10.1007/s00190-019-01335-w

Write a comment

Comments: 2
  • #1

    Bauer (Friday, 18 September 2020 10:03)

    Finally! So nice to see you again

  • #2

    Manoj Deo (Friday, 18 September 2020 20:35)

    Thanks for the update. This is a great service for users out there.