Assessment of Precise Point Positioning (PPP) for Enhancing Georeferencing Accuracy in UAV Photogrammetry: Accuracy Analysis and Applicability

Precise Point Positioning (PPP) offers a single-receiver alternative to differential GNSS for UAV direct georeferencing, although its kinematic application remains constrained by convergence time and the robustness of filter-based estimation under motion. This study investigates post-processed dual-frequency kinematic PPP in float mode under two observation configurations, namely GPS-only and combined GPS+BeiDou, and further evaluates GPS-only PPP with ambiguity resolution (PPP-AR). Performance is assessed against a tightly coupled GNSS/INS PPK reference trajectory. Convergence is quantified using a time-to-sustained-compliance criterion, defined as the first epoch at which both horizontal and vertical errors remain continuously within mapping-grade thresholds over a five-minute interval. The results indicate that PPP-float converges in 18.49 min with GPS-only observations, whereas convergence degrades to 32.37 min when BeiDou is included, suggesting delayed stabilization associated with the estimation of additional inter-system parameters and constellation-dependent residual biases. Since the JoUAV CW-15II receiver used in this study does not support newer BDS-3 signals, the poorer convergence of the combined solution is plausibly linked to the dominant contribution of BeiDou GEO and legacy BDS-2 satellites, whose observations and correction products are generally of lower quality than those of GPS. This finding suggests that tracking BDS-3 with upgraded hardware may improve future GPS+BeiDou PPP-float performance. In contrast, GPS-only PPP-AR converges in 14.92 min and improves post-convergence accuracy, reducing the 3D RMSE from approximately 0.086 m to 0.041 m. As PPP-float and PPP-AR were processed using different software engines, the comparison is presented at the workflow level to support practical mission planning.