Evaluating survey instruments and methods in a steep channel

Abstract

Methods for surveying and analyzing channel bed topography commonly lack a rigorous characterization of their appropriateness for project objectives. We compare four survey methods: a hand level, two different methods of surveying with a laser rangefinder, and a real-time kinematic GNSS (RTK-GNSS) to explore their accuracy in determining channel bed slope and roughness for a study reach in a small, dry, steep channel. Additionally, we evaluate the variability among four operators for each survey technique. Two methods of calculating reach slope were computed: a regression on the channel profile and a calculation using only survey endpoints. Using data from the RTK-GNSS as our accuracy reference, the hand level and two-person laser rangefinder surveying systems performed with high accuracy (<5% error in estimating slope, <10% error in estimating roughness), while the one-person laser rangefinder survey system performed with considerably lower accuracy (up to 54% error in roughness and slope). Variability between operators was found to be very low (coefficients of variation ranged from 0.001 to 0.046) for all survey systems except the one-person laser rangefinder system, suggesting that survey data collected by different operators can be validly compared. Due to reach-scale concavity, calculating slope using a regression produced significantly different values than those obtained by using only survey endpoints, suggesting that caution must be taken in choosing the most appropriate method of calculating slope for a given project objective. We present recommendations for choosing appropriate survey and analysis methods to accomplish various surveying objectives.