Static and dynamic hydro-acoustic assessment of sediment magnitudes in a Brazilian reservoir

Abstract

The importance of reservoirs around the world is rising, not only due to the fact that more and more reservoirs are build, but also because the resulting dependence for electricity and water supply is increasing. Nearly all reservoirs suffer from sedimentation problems and therefore the quantification of sediment accumulation plays an important role for the reservoir management. Several methods for the estimation of sediment thickness in reservoirs are available. One option is the gravity and vibrational coring. It has the disadvantage of relatively high handling and working efforts in order to recover sediment cores with sufficient depth. This approach is also very limited in regards to spatial information, since the cores only contain sediment information for the current position. Another option is the Topographic Differencing, where the difference between the height information of the pre- and post-impoundment lakebed is calculated. In many cases, the quality of the results is limited by the poor quality of pre-impoundment maps. This can be due to missing information about the map coordinate system or transformation and the insufficient resolution caused by an inappropriate spatial scale of the map. As a third, and most promising option the Acoustic Estimation of the sediment thickness was conducted. Data was obtained during two campaigns, 2012 and 2013 on the Capivari reservoir in south-east Parana state, Brazil. A 200 kHz frequency was used to determine the sediment water interface and a 38 kHz frequency for the detection of the lower boundary of the soft sediment overlay. Boundary detection and data processing was conducted using the Sonar5Pro software. As the reflected strength (Ts) of the echo signal does not reach the same level for all sediment types and locations, a determined threshold for the detection of the interface could not be derived. Instead of a fixed threshold for the lower boundary detection, the echogram data is visually analyzed and the highest reflection pixel cluster is marked as the sediment layer interface. Sediment thickness was assessed at 18 static positions as well as during a dynamic survey, which covered nearly the entire reservoir. For sediment overlays up to 60 cm a good accordance between the hydro-acoustic and the sediment information was found. Penetration depths of the gravity coring resemble a direct relation to the assessed sediment magnitude (r = 0.7); the higher the sediment overlay the deeper penetrated the corer. Interpolated sediment thicknesses were validated with the static hydro-acoustic measurements. Data shows that the driven survey produced thickness values, which match the results from the static measurements (r = 0.92). The semi-automated detection of the lower sediment overlay boundary proves to be able to reproduce local heterogeneities of sediment thickness, which is shown to be strongly dependent on the inclination of the lake bed.