[1] Large-scale particle image velocimetry (LSPIV) is a nonintrusive approach to measure velocities at the free surface of a water body. The raw LSPIV results are instantaneous water surface velocity fields, spanning flow areas up to hundreds of square meters. Measurements conducted in typical conditions in conjunction with appropriate selections of parameters for image processing resulted in mean velocity errors of less than 3.5%. The current article reviews the background of LSPIV and the work of three research teams spanning over a decade. Implementation examples using various LSPIV configurations are then described to illustrate the capability of the technique to characterize spatially distributed two-and three-dimensional flow kinematic features that can be related to important morphologic and hydrodynamic aspects of natural rivers. Finally, results and a critique of research methods are discussed to encourage LSPIV use and to improve its capabilities to collect field data needed to better understand complex geomorphic, hydrologic, and ecologic river processes and interactions under normal and extreme conditions. Citation: Muste, M., I. Fujita, and A. Hauet (2008), Large-scale particle image velocimetry for measurements in riverine environments, Water Resour. Res., 44, W00D19,

Flash-floods that occur in Mediterranean regions result in significant casualties and economic impacts. Remote imagebased techniques such as Large-Scale Particle Image Velocimetry (LSPIV) offer an opportunity to improve the accuracy of flow rate measurements during such events, by measuring the surface flow velocities. During recent floods of the Ardèche river, LSPIV performance tests were conducted at the Sauze-Saint-Martin gauging station without adding tracers. The rating curve is well documented, with gauged discharge ranging from 4.8 m 3 s −1 to 2700 m 3 s −1 , i.e., mean velocity from 0.02 m s −1 to 2.9 m s −1. Mobile LSPIV measurements were carried out using a telescopic mast with a remotely controlled platform equipped with a video camera. Also, LSPIV measurements were performed using the images recorded by a fixed camera. A specific attention was paid to the hydraulic assumptions made for computing the river discharge from the LSPIV surface velocity measurements. Simple solutions for interpolating and extrapolating missing or poor-quality velocity measurements, especially in the image far-field, were applied. Theoretical considerations on the depth-average velocity to surface velocity ratio (or velocity coefficient) variability supported the analysis of velocity profiles established from available gauging datasets, from which a velocity coefficient value of 0.90 (standard deviation 0.05) was derived. For a discharge of 300 m 3 s −1 , LSPIV velocities throughout the river crosssection were found to be in good agreement (±10%) with concurrent measurements by Doppler profiler (ADCP). For discharges ranging from 300 to 2500 m 3 s −1 , LSPIV discharges usually were in acceptable agreement (< 20%) with the rating curve. Detrimental image conditions or flow unsteadiness during the image sampling period led to larger deviations ranging 30-80%. The compared performances of the fixed and mobile LSPIV systems evidenced that for LSPIV stations, sampling images in isolated series (or bursts) is a better strategy than in pairs evenly distributed in time.

Large Scale Particle Image Velocimetry (LS-PIV) is used to measure the surface flow velocities in a mountain stream during high flow conditions due to a reservoir release. A complete installation including video acquisition from a mobile elevated viewpoint and artificial flow seeding has been developed and implemented. The LS-PIV method was adapted in order to take into account the specific constraints of these high flow conditions. Using a usual LS-PIV data processing, significant variations of the water surface elevation were taken into consideration in the image rectification. An intensity threshold was applied to focus on artificial tracers without considering stationary waves and sun reflections on the flow surface. A site-specific float coefficient of 0.79 based on measured vertical velocity profiles was used to convert surface velocities into depth-averaged velocities. Comparison between LS-PIV assessments and 2Dh numerical calculations with the code Rubar20 allows verification and extrapolation of LS-PIV data. LS-PIV velocity measurements permit to assess discharges over the whole high flow event in agreement with leaded current-meter measurements performed at a downstream bridge

This paper investigates the potential of fast flood discharge measurements conducted with a mobile LSPIV device. LSPIV discharge measurements were performed during two hydrological events on the Arc River, a gravel-bed river in the French Alps: a flood greater than the 10-year return period flood in May, 2008, and a reservoir flushing release in June, 2009. The mobile LSPIV device consists of a telescopic mast with a remotely controlled platform equipped with a video camera. The digital video camera acquired sequences of images of the surface flow velocities. Ground Reference Points (GRPs) were positioned using a total station, for further geometrical correction of the images. During the flood peak, surface flow velocities up to 7 m/s and large floating objects prevented any kind of intrusive flow measurements. For the computation of discharge, the velocity coefficient was derived from available vertical velocity profiles measured by current meter. The obtained value range (0.72e0.79) is consistent with previous observations at this site and smaller than the usual default value (0.85) or values observed for deeper river sections (0.90 typically). Practical recommendations are drawn. Estimating stream discharge in high flow conditions from LSPIV measurements entails a complex measurement process since many parameters (water level, surface velocities, bathymetry, velocity coefficient, etc.) are affected by uncertainties and can change during the experiment. Sensitivity tests, comparisons and theoretical considerations are reported to assess the dominant sources of error in such measurements. The multiplicative error induced by the velocity coefficient was confirmed to be a major source of error compared with estimated errors due to water level uncertainty, free-surface deformations, number of image pairs, absence or presence of artificial tracers, and cross-section bathymetry profiles. All these errors are estimated to range from 1% to 5% whereas the velocity coefficient variability may be 10%e15% according to the site and the flow characteristics. The analysis of 36 LSPIV sequences during both events allowed the assessment of the flood discharges with an overall uncertainty less than 10%. A simple hydraulic law based on the geometry of the three sills of the Pontamafrey gauging station was proposed instead of the existing curve that is fitted on available gauging data. The high flow LSPIV discharge measurements indicated that this new curve is more accurate for high discharges since they are evenly distributed in a AE10% interval around it. These results demonstrate the interest of the remote stream gauging techniques together with hydraulic analysis for improving stageedischarge relationships and reducing uncertainties associated with fast flood discharges.

Movies taken by witnesses of extreme ood events are increasingly available on video sharing websites. They potentially provide highly valuable information on ow velocities and hydraulic processes that can help improve the post-ood determination of discharges in streams and ooded areas. We investigated the troubles and potential of applying the now mature LSPIV technique to such ood movies that are recorded under non-ideal conditions. Processing was performed using user-friendly, free software only, such as Fudaa-LSPIV. Typical issues related to the image processing and to the hydrological analysis are illustrated using a selected example of a pulsed ash-ood ow lmed in a mountainous torrent. Simple corrections for lens distortion (sheye) and limited incoherent camera movement (shake) were successfully applied and the related errors were reduced to a few percents. Testing the dierent image resolution levels oered by YouTube showed that the dierence in time-averaged longitudinal velocity was less than 5% compared to full resolution. A limited number of GRPs, typically 10, is required but they must be adequately distributed around the area of interest. The indirect determination of the water level is the main source of uncertainty in the results, usually much more than errors due to the longitudinal slope and waviness of the ow free-surface. The image-based method yielded direct discharge estimates of the base ow between pulses, of the pulse waves, and of the time-averaged ow over a movie sequence including a series of 5 pulses. A comparison with traditional indirect determination methods showed that the criticaldepth method may produce signicantly biased results for such a fast, unsteady ow, while the 1 Author-produced version of the article published in Hydrological Processes (2016), Volume 30, Issue 1, p 90-105The original publication is available at http://onlinelibrary.wiley.com DOI: 10.1002/hyp.10532 slope-area method seems to be more robust but would overestimate the time-averaged ow rate if applied to the high-water marks of a pulsed ow.