Date Awarded


Document Type


Degree Name

Master of Science (M.Sc.)


Virginia Institute of Marine Science


The Sensor Insertion System (SIS) located at the US Army Corps Field Research Facility (FRF) in Duck, NC, allows closely spaced measurements of along-shore velocity and suspended sediment concentration to be collected across the entire surf zone during storms. The SIS infers suspended sand concentration from backscatter off of suspended material using optical backscatter (OBS) sensors. Unfortunately, OBS response is highly sensitive to grain size, including both differences between sand and mud and variations in size within the sand range. As part of this study, ground-truthing measurements were added to the SIS during a storm in October 1997, including measurements of suspended sand concentration via pump sampling, by laser in situ scattering and transmissometery (LISST) and by acoustic backscatter (ABS). The major objectives of this study were to use the resulting data to address the OBS's sensitivity to grain size and suggest corrective measures.

The first objective was to test the "cutoff" method for the removal of background turbidity due to fine sediment. The study found the cutoff value, defined by some lowest percentile of the OBS response during a given burst, to be proportional to the pumped concentration of suspended particles smaller than 63 microns in diameter. The best choice of cutoff value (1% to 5%) was relatively insensitive to the precise cutoff percentage, indicating the 5% cutoff value currently used by the FRF works well. Addressing the second objective, OBS response after the removal of fines was found to be consistent with pumped sand concentration as long as corrections were made for (i) the varying size of suspended sand, (ii) the precise time of pump sampling and (iii) for apparent noise in the OBS records. Correction for the smaller size of suspended sand (average size 120 microns) relative to that used during calibration (average size 230 microns) decreased OBS estimates of sand concentration by about 42%. Accounting for noise decreased OBS estimates of sand concentration by as much as 80%. Addressing the third objective, LISST and ABS measurements of suspended sand were compared with standard laboratory rapid sand analysis (RSA). Burst-Averaged D50 sand grain size estimated with the LISST was found to have a reasonable correlation with RSA D50 sand grain size estimations. Suspended sand concentrations estimated by the LISST, by the ABS 2.5 MHz transponder and by weighing of pumped samples agreed well.

The fourth objective was development of a model for determining sand size from OBS and current meters without relying on ABS, LISST or pump sampling. According to established theory, the ratio of the sand settling velocity (w) to the hydrodynamic shear velocity (u*) can be estimated from the slope of a log-log plot of the burst-averaged concentration profile versus height above the bed. Independent estimation of u* should then give w from the OBS profile and therefore the suspended sand grain size. This relationship was modified for application to OBS data by accounting for (i) a normal distribution of sand grain size and (ii) the OBS's inverse sensitivity to grain size. Theoretical models for u* due to waves and currents were then applied but did not adequately predict the measured sand grain size. A best-fit u* was derived from the observed OBS profiles by combining the observed ratio of w/u* with the directly measured grain size. The best-fit u* was then plotted against current velocity (uc) and showed a change in behavior of u* for uc > 63 cm/sec. Two multiple regressions were performed between the best-fit u* and various parameters routinely measured by the FRF. The first regression, for stations with uc > 63 cm/sec, found uc, water depth and wave height to be significant predictors of u*. For uc < 63 cm/sec, wave orbital velocity, current velocity and bottom roughness were significant. Finally, addressing the last objective, a step-by-step method to correct OBS records for estimating sand concentration was presented which included (1) removal of the effects of background turbidity, (2) removal of instrument noise, (3) estimation of sand size from the shape of the OBS profile, and (4) adjustment of measured OBS concentration based on the estimated sand grain size.



© The Author

Included in

Oceanography Commons