Estimating Error in Geophysical Sub-bottom Profiling

摘要

Two complimentary approaches, statistical and analytical, are considered when estimating the accuracy of surveys that are carried out to determine quantities of sediments for dredging. Early surveys for such estimation are done by sampling on a grid system that gave little indication of the variability of such sediments between the sampling points so that only rough estimates could be made of the quantities. This could result in substantial cost overestimation in the order of many millions of Euros. Continuous geophysical reflection surveys were made at numerous locations in the Netherlands as well as case studies in the harbours of Bergen and Gothenburg. The accuracy of seismic-reflection surveys (also referred to as sub-bottom profiling) is a function of the frequencies of the acoustic signal (ω), the frequency at which values are obtained from the acoustic profiles (sampling frequency f_s), position, and interpolation between traverse lines. Other factors play a role such as interpretation, poor, virtual or migrated reflectors, sediment mobility, truthing and, ultimately, objectives of the sediment survey. The variability of the top sediment could be determined along a traverse line and is compared with the resulting variability obtained from interpreted contouring between the traverse lines. An example is presented for a survey carried out in the Alvsborg basin of Gothenburg Harbour. Top sediment isopach variability corresponds to the average variability of the profiles. Coefficient of variation (CoV) is used to quantify variability. Sampling frequency and sedimentation energy levels both have an effect on variabilit