Authors: Evan M Lindroth*, Geography, University of Illinois, Urbana-Champaign, Bruce L Rhoads, Geography, University of Illinois, Urbana-Champaign, Quinn W Lewis, Geography, University of Illinois, Urbana-Champaign
Topics: Water Resources and Hydrology, Quantitative Methods
Keywords: LSPIV, stream gauging, UAS, river discharge
Session Type: Paper
Start / End Time: 5:20 PM / 7:00 PM
Room: Balcony K, Marriott, River Tower Elevators, 4th Floor
Presentation File: No File Uploaded
Large-scale particle image velocimetry has been proposed as a low-cost method to rapidly measure river discharge. This method is especially useful for measuring fast-changing or otherwise dangerous flows. Despite its potential, few, if any, studies have thoroughly evaluated the accuracy of LSPIV discharge measurements using different image-based measurement platforms. This paper evaluates the accuracy of LSPIV discharge measurements derived from imagery of the water surface obtained from a fixed tripod and an unmanned aerial system (UAS), achieved through a detailed comparison among LSPIV, hydroacoustic, current-meter, and gauged discharge measurements at multiple field sites. Thirteen LSPIV and in-channel discharge measurements were obtained for a wide range of hydrodynamic conditions. Additional measurements were collected with the purpose of evaluating the effects of camera setup and LSPIV processing on resultant velocity and discharge. Discharge estimates were reasonably accurate compared to reference measurement techniques. Mean error between LSPIV and hydroacoustic measurements was 10%. Mean error between LSPIV and mechanical measurements was 19%, and mean error between hydroacoustic and mechanical measurements was 20%. The results of this study indicate that if important minimum requirements are met, UAS LSPIV performs no worse than fixed LSPIV and produces discharge estimates with a similar accuracy and repeatability as in-channel measurement methods. LSPIV setups using either UAS or fixed platforms can be rapidly set up by small teams even with no prior site preparation. The accuracy, repeatability, and low cost of this method should appeal to a wide range of end users interested in obtaining data on river flow.