PPP not converging? Check observations, products, antennas and frames
If a PPP solution does not converge, first check continuous carrier-phase observations, compatible orbit and clock products, antenna metadata and processing mode. Then inspect residuals and parameter time series. More observation time cannot correct the wrong antenna height or make inconsistent products compatible.
What convergence means
PPP estimates coordinates and nuisance parameters from undifferenced observations using precise satellite information. Early in a session, clocks, atmospheric terms and ambiguities may be weakly separated. Convergence describes the stabilisation of the solution under the selected model. It has no universal duration that applies to every receiver, signal environment and PPP service.
Use the symptom to choose the next check
| Symptom | Check first |
|---|---|
| Repeated restarts or coordinate jumps | Phase gaps, cycle slips, clock events and changes in satellite availability |
| Missing satellites or long rejected intervals | Observation codes and the time and constellation coverage of precise products |
| Height is stable but offset | Antenna type, radome, height convention, tide convention and vertical reference |
| Smooth coordinates disagree with known control | Frame realisation, coordinate epoch and the reference coordinates used for comparison |
| Static mode fails on a moving platform | Use a kinematic model if the service supports the intended observations and dynamics |
Check the actual observation arcs
Inspect epoch spacing, code and phase availability, and the continuity of each satellite’s tracked signals. A long file with frequent phase interruptions is not equivalent to an uninterrupted observation session of the same duration. Run QC before positioning and retain its report with the processing results.
- Check whether a frequency needed by the processing strategy is missing.
- Look for session-wide events affecting many satellites together.
- Confirm that the antenna remained stationary if the job is configured as static.
Sources: ESA Navipedia: Precise Point Positioning · IGS: formats and standards
Audit the orbit, clock and calibration inputs
Use the product combinations supported by the PPP engine. Verify timestamps and satellite coverage in the files rather than relying on a download-success message. Match receiver and satellite antenna corrections to the processing conventions, and check any bias products required by the ambiguity-resolution strategy.
- For a session spanning two days, check product coverage across the day boundary.
- Record the analysis centre, product series, filenames and processing version.
- Do not assume a different product latency category is automatically better if it is incompatible with the chosen model.
Sources: IGS: GNSS products · IGS: formats and standards
Test one change at a time
Keep one baseline run, change a single suspect input and compare the resulting residuals, accepted observations and coordinate time series. If the issue is an antenna or frame mismatch, correct that first. Only then evaluate whether a longer continuous session improves the result. An independent check against reliable control is stronger evidence than a small formal uncertainty alone.
Keep a reproducible support package
- Original observation file, station log and antenna height convention.
- Run settings, product filenames, warnings and processing report.
- The coordinate system, frame realisation, epoch and height type expected in the deliverable.
- A precise description of the observed failure, including its time range.
References and methodology
Technical definitions follow the sources below. Worked examples are illustrative; product-specific thresholds are identified as PosFlow settings. For corrections, contact PosFlow with the article URL and the relevant specification.