Case Study of Precision of GPS Differential Correction Strategies: Influence on aDcp Velocity and Discharge Estimates

The precision of four differential global positioning systems (DGPS) was evaluated in the context of fluvial water velocity and discharge measurement. DGPS is used to resolve water velocities measured with an acoustic Doppler current profiler (aDcp) into earth coordinates if bottom tracking is unavailable. The DGPS systems assessed were: (1) the dual frequency real time kinematic (RTKL1L2); (2) the single frequency real time kinematic (RTKL1); (3) the code-phase Canadian Coast Guard (CG); and (4) the code-phase Wide Area Augmentation System (WAAS). Repeat discharge surveys (n=22) were conducted at a transect of the Gatineau River, Canada, simultaneously collecting bottom track boat velocity (vBT) and boat velocity from all four DGPS (vDGPS). The mean absolute single ping differences between vBT and vDGPS were 3.1 (RTKL1L2), 3.2 (RTKL1), 8.9 (CG), and 9.8 cm/s (WAAS). Errors were observed more often near channel margins, presumably due to obstruction and multipath associated with riverbank vegetation and buildings. DGPS velocity errors were random, and a large number of DGPS positions were utilized across the section to record discharge. Accordingly, errors in discharge were relatively small, with maximum percentage differences between single transect QBT and QDGPS of 0.9 (RTKL1L2), 1.0 (RTKL1), 2.4 (CG), and 3.1% (WAAS). Simulations suggest large discharge errors (up to 51%) are possible under low sampling intensity (20 pings) and small channel velocity relative to average vDGPS error (ratio of 1).