# Impact of Satellite Clock Corrections and Different Precise Products on GPS and Galileo Precise Point Positioning Performance

## TL;DR

Shorter satellite clock intervals and better clock products improve GPS and Galileo positioning accuracy and speed.

## Contribution

This study evaluates the impact of different satellite clock products and sampling rates on GPS and Galileo PPP performance.

## Key findings

- Shorter clock sampling (30 s) improves PPP accuracy and reduces convergence time.
- Final and operational products outperform rapid and ultra-rapid ones, especially for GPS.
- Linear interpolation benefits Galileo but harms GPS when using 5 min clocks.

## Abstract

What are the main findings?
Shorter satellite clock sampling (30 s vs. 5 min) yields more accurate PPP and faster convergence time. Final and operational (OPS) products outperform rapid and MGEX; ultra-rapid—especially the predicted half—are the weakest.Linear interpolation of clock corrections is constellation-dependent: with 5 min clocks, it benefits Galileo but often harms GPS. In GPS+Galileo, the effects largely cancel, with a slight GPS-driven tendency. With 30 s clocks, interpolation is negligible.

Shorter satellite clock sampling (30 s vs. 5 min) yields more accurate PPP and faster convergence time. Final and operational (OPS) products outperform rapid and MGEX; ultra-rapid—especially the predicted half—are the weakest.

Linear interpolation of clock corrections is constellation-dependent: with 5 min clocks, it benefits Galileo but often harms GPS. In GPS+Galileo, the effects largely cancel, with a slight GPS-driven tendency. With 30 s clocks, interpolation is negligible.

What is the implication of the main finding?
For PPP with 30 s data, use dedicated, dense CLK products from Final/OPS and combine GPS+Galileo; treat ultra-rapid—particularly the predicted segment—cautiously.If only 5 min SP3 clocks are available, apply linear interpolation (CLOCK1) for Galileo and nearest-epoch sampling (CLOCK0) for GPS; however, shortening the clock interval matters more than the interpolation strategy.

For PPP with 30 s data, use dedicated, dense CLK products from Final/OPS and combine GPS+Galileo; treat ultra-rapid—particularly the predicted segment—cautiously.

If only 5 min SP3 clocks are available, apply linear interpolation (CLOCK1) for Galileo and nearest-epoch sampling (CLOCK0) for GPS; however, shortening the clock interval matters more than the interpolation strategy.

This study assesses how satellite clock products affect Precise Point Positioning (PPP) for GPS, Galileo, and GPS+Galileo. Multi-GNSS data at 30 s were processed for 12 global IGS stations over one week in 2025, with each day split into eight independent three-hour sessions. SP3 clocks (ORB, 5 min) were compared with dedicated CLKs (CLO, 5 s, 30 s, 5 min) across final (FIN), rapid (RAP), and ultra-rapid (ULT; observed/predicted) product lines from multiple analysis centers. Two timing strategies were tested: nearest-epoch sampling (CLOCK0) and linear interpolation (CLOCK1). CLO consistently delivered the lowest 2D/3D errors and the fastest convergence. ORB degraded accuracy by a few millimeters and extended convergence by ~5–10 min, most notably for GPS. With 5 min clocks, CLOCK1 yielded small gains for Galileo but often hurt GPS; with 30 s clocks, interpolation was immaterial; 5 s clocks offered no measurable benefit. FIN outperformed RAP; OPS slightly outperformed MGEX; ESA/GFZ ranked highest. ULT solutions were weaker, especially in the predicted half. Zenith tropospheric delay (ZTD) biases were negligible; variance was smallest for GPS+Galileo with CLO (~7–10 mm), increased by ~1–2 mm with ORB, and was largest in ULT. Dense, high-quality clock products remain essential for reliable PPP.