3rd HyMeX workshop 1-4 June 2009 Heraklion (Gournes), Crete-Greece
Flash-flood monitoring by image analysis in the Ardèche river catchment, France: validation tests and set-up of a distributed observation networkIsabelle Braud (Cemagref UR HHLY); Jérôme Le Coz, Guillaume Dramais, Alexandre Hauet
Flash-floods that occur in the southern Mediterranean region of France result in significant casualties and economic impacts. The understanding of flood generation and propagation processes requires reliable streamflow estimates throughout the river network. Unfortunately, the stage-discharge relationships (rating curves) empirically established at existing gauging stations often must be extrapolated to high flow rates, resulting in large uncertainty levels due to the lack of direct discharge measurements. Remote techniques such as image analysis (LSPIV) offer an opportunity to improve the accuracy of high flow rate measurements, by measuring the surface flow velocities.
During recent floods, discharge measurement tests using the LSPIV technique were conducted at the Sauze-Saint-Martin gauging station, near the outlet of the Ardèche river catchment (2240 km², mean annual discharge 63.4 m3/s, Q5=2530 m3/s). Mobile LSPIV measurements were carried out using a telescopic mast with a remotely controlled platform equipped with a video camera. Also, fixed LSPIV measurements were performed using the images recorded by a camera installed at the station. At Sauze-Saint-Martin, the rating curve is well documented, with streamflow gaugings up to 2700 m3/s. A float coefficient value of 0.90 was derived from the analysis of velocity profiles established from available gauging datasets.
During a small flood (400 m3/s), LSPIV tests were compared to concurrent discharge measurements by Doppler profiler (aDcp). Depth-averaged aDcp velocities and LSPIV velocities corrected by the float coefficient were found to be in good agreement (<10%) throughout the river cross-section. Both aDcp and LSPIV discharges fell within 4% of the values given by the rating curve. During larger floods (900 to 2500 m3/s), discharges yielded by the fixed LSPIV station were found to be in acceptable agreement (<20%) with the rating curve. Tests were also conducted at other cross-sections throughout the Ardèche catchment, planned to be equipped with fixed cameras. Main research perspectives aim at improving the computation of LSPIV discharge (float coefficients, hydraulic modelling), and of associated uncertainty.