Implementation of comparative detection approaches for the accurate assessment of sediment thickness and sediment volume in the Passaúna Reservoir

Siltation has significant economic and social impacts as it directly reduces the useable amount of water in reservoirs. Giving a solution to the issue of sedimentation is a complicated task and maybe one of the most important engineering and environmental challenges of the 21^st century. The deposited volume and the distribution pattern of the sediment are often unknown and not easy to assess. The sedimentation process is highly dynamic, initially due to the hydrological conditions of the incoming rivers, but also due to common internal phenomena like resuspension or density currents. Sediment remediation measures such as mechanical sediment removal or flushing are planned based on the sediment thickness distribution and the overall sediment volume/mass. Often, the sediment thickness is calculated through topographic differencing between the pre-impoundment reservoir lake bottom and the actual lake bottom. However, it is common that the previous depth distribution map is not available or in insufficient quality. In this regard, alternative measurement techniques have to be taken into consideration. In this study, we assessed the best possible approach depending on the characteristics of the sediment and of the reservoir. We combined three different acoustic systems (a multibeam, a sub-bottom profiler, and a single beam dual frequency system) with sediment coring and dynamic free fall penetrometer measurements for an improved assessment of sediment stock and sediment distribution in the Passaúna Reservoir. Our results showed that topographic differencing could not be applied, as the data for the pre-impoundment lake bottom was insufficiently accurate. The parametric sub-bottom profiler could detect the sediment thickness in high accuracy, but significant limitations were recorded in areas with high gas contents. The dual-frequency echosounder derived the sediment thickness with a normalized mean absolute error of 56% due to the high volumetric gas content in the sediment. The dynamic free-fall penetrometer showed satisfying results compared to the other systems. The normalized mean absolute error was 22%, and sediment thickness could be detected in areas with up to 1.8 m of sediment. Sediment coring is also a reliable technique for sediment thickness determination. However, the results showed that if only traditional coring devices are used (gravity corer), the limited penetration depth of the equipment combined with sampling disturbances often prevent a correct assessment of the sediment thickness. The overall results of this study can help for an improved decision-making regarding reservoir management. The accurate assessment of sediment volume and distribution can reduce costs for sediment removal and assist in having a precise overview of the reservoir lifetime.