USGS Home
Contact USGS
Search USGS

U.S. Geological Survey
Open-File Report 91-375-A
Note: This is the text portion of the report. The appendices are not included.
See the Full Report online at the USGS Publications Warehouse:
http://pubs.er.usgs.gov/usgspubs/ofr/ofr91375A

Station Locations and Grain-Size Analysis of Surficial Sediment Samples collected on the Continental Shelf, Gulf of the Farallones during Cruise F2-89-NC, January 1989

N. M. Maher, H. A. Karl, J. L. Chin and W. C. Schwab

INTRODUCTION

The U.S. Geological Survey began a major geologic and oceanographic study of the Gulf of the Farallones in 1989. This investigation, the first of several planned adjacent to major population centers, was designed to establish a good scientific data base on a segment of continental shelf adjacent to the San Francisco Bay area so as to evaluate and monitor human impact on the marine environment. Two cruises were conducted in January 1989 on the continental shelf between Cordell Bank and Half Moon Bay east of the Farallon Islands (Chin et al., 1989). Reconnaissance side-scan sonar and high-resolution seismic-reflection surveys were conducted along a rectilinear grid of tracks spaced nominally 4 km apart. This report describes the collection and analysis of 268 surficial sediment samples collected during the second cruise, F2-89-NC.

DATA COLLECTION

A regional grid of 97 samples of surficial sediment was collected with a Soutar Van Veen sampler at the intersections of the reconnaissance tracklines and, at other sites of interest, at intervals of about 4 km (Figure 1). A denser grid of 171 samples spaced nominally 1 km apart was collected just east of the Farallon Islands (Figure 1 and Figure 2).

Three systems were used to navigate the ship during the sampling: (1) Global Positioning System (GPS), (2) LORAN-C, and (3) a shore-based transponder net. The primary system used for real-time positioning was chosen either manually by the navigator or automatically by the computer. Positional accuracy varied from about 100 m using GPS and LORAN-C to a few meters using the shore-based transponder net.

LABORATORY ANALYSIS

Grain-size Analysis

The following procedures used to determine grain-size distribution of the sediment samples were standardized by the U.S. Geological Survey Marine Geology laboratory, Menlo Park. Similar methods are described in Carver (1971) and Theide and others (1976).

Oxidation of of organic matter

Five to ten gram portions of sample were placed in a beaker with 200 ml of distilled water and 5 ml of 30% hydrogen peroxide. The samples were stirred to fully disaggregate the sample and heated overnight at 70-80° C to drive off excess hydrogen peroxide following oxidation of organic matter.

Removal of soluble salts

Following oxidation, the samples were washed into plastic 250 ml bottles and centrifuged at about 1700 rpm for 45 minutes. The water and soluble salts were decanted, replaced with distilled water, centrifuged a second time, and again decanted.

Size segregation

After removal of soluble salts, the samples were wet sieved with distilled water into three size fractions: Gravel (>2 mm), sand (63 - 2 mm), and mud (silt + clay) (<63 ). The gravel and sand fractions were transferred to pre-weighed evaporating dishes, dried, and reweighed. The weight of the sample was then calculated by subtracting the weight of the dish. The silt + clay fractions were transferred into graduated cylinders and filled to the 1 liter mark with distilled water. Five ml of 10% sodium hexametaphosphate solution was added to disperse clay and clay size particles and prevent flocculation.

Determining the weight and size distribution of the silt + clay (<63 ) fraction

A representative 20 ml aliquot was pipetted from each 1 liter graduated cylinder, transferred to a preweighed aluminum weighing boat, oven dried and reweighed. The total weight of the <63 fraction was calculated by multiplying that weight by 50 and subtracting out the sodium hexametaphosphate weight. Weight percents for half-phi intervals from 4.0 to 9.0 (phi sizes are converted to millimeters as explained below; see also Table 1) were determined using a hydrophotometer (Jordan et al., 1977).

Determining the size distribution of the sand fraction

A Rapid Sediment Analyzer (RSA), commonly known as a settling tube, was used to measure weight percents for half-phi intervals from -1.0 to 4.5 phi. In the RSA, the sand grains are released at the top of a 2 m water column and settle to a weighing platform . The cumulative weight vs time is recorded on a chart recorder from which the size distribution can be measured using a calibrated overlay. Theory, equipment, and techniques employed are described in detail by Theide and others (1976).

Calculation of grain-size statistics and parameters

Statistics of the grain-size distribution were computer calculated with a U.S. Geological Survey Marine Geology grain-size program (McHendrie 1988). Grain-size statistical parameters and graphic representations are given in phi units. The phi unit () is a logarithmic transformation of millimeters into whole integers, according to the formula:

where d = grain diameter in millimeters.

The parameters calculated for these analyses include:

1. "median" - corresponds to the 50 percentile on a cumulative curve, where half the particles by weight are larger and half are smaller than the median. This parameter is measured in phi units.
2. "mean" - is the average grain-size. Several formulas are used in calculating the mean. The most inclusive graphically derived value is that given by Folk (1968):
   where 16, 50, and 84 represent the size at 16, 50, and 84 percent of the sample by weight. Mean is also measured in phi units and is the most widely compared parameter.
3. "sorting" - is a method of measuring the grain-size variation of a sample by encompassing the largest parts of the size distribution as measured from a cumulative curve. Folk (1968) introduced the "inclusive graphic standard deviation", that is calculated as follows:
   
   
   
   where 84, 16, 95, and 5 represent the phi values at 84, 16, 95, and 5 percentiles. Folk (1968) presented a verbal classification scale for sorting (Table 2): <0.350: very well sorted; 0.35-0.500: well sorted; 0.5-0.710: moderately well sorted; 0.71-1.00: moderately sorted; 1.00-2.00: poorly sorted; 2.00-4.00: very poorly sorted; and, >4.00: extremely poorly sorted.
4. "skewness" - measures the degree to which a cumulative curve approaches symmetry. Two samples may have the same average grain size and sorting but may be quite different to their degrees of symmetry. Folk's "inclusive graphic skewness" (1968) is determined by the equation:
   
   
   
   where the phi values represent the same percentages as those for sorting. This formula includes a measure of the "tails" of the cumulative curve as well as the central portion. Other methods for determining skewness, notably those by Inman (1952) and Trask (1950), do not measure the tails of the curve. Symmetrical curves have a skewness equal to 0.00; those with a large proportion of fine material are positively skewed; those with a large proportion of coarse material are negatively skewed. A verbal classification for skewness suggested by Folk (1968) includes (Table 2): from +0.10 to -0.10 as nearly symmetrical; -0.10 to -0.30 as coarse-skewed; and, -0.30 to -1.00 as strongly coarse-skewed.
5. "kurtosis" - is a measure of "peakedness" in a curve. Folk's (1968) formula for kurtosis is:
   
   
   
   where the phi values represent the same percentages as those for sorting. A normal Gaussian distribution has a kurtosis of 1.00 which is a curve with the sorting in the tails equal to the sorting in the central portion. If a sample curve is better sorted in the central part than in the tails, the curve is said to be excessively peaked, or leptokurtic; if the sample curve is better sorted in the tails than in the central portion, the curve is flat peaked or platykurtic. For normal curves = 1.00, leptokurtic curves have >1.00, and platykurtic curves have <1.00.

Method of Moments

All of the above statistical parameters can be calculated using the method of moments. This method gives a more rigorous treatment of the sediment characteristics. The computer program used for the sample analyses in this study performed the necessary calculations for parameter determination. The first moment measure corresponds to the mean, the second to the standard deviation, the third to the skewness, and the fourth to the kurtosis.

All data derived from these analysis are shown in Appendix I and II. Grain-size parameters calculated using the method of moments and graphically derived values are both reported.

PRELIMINARY RESULTS AND INTERPRETATION

The distribution of surficial sediment textures based on the regional grid of 97 stations suggests that depositional processes in the Gulf of the Farallones are complex. A 20-km wide corridor of sand extends westerly from the Golden Gate to the Farallon Islands (Figure 3). Sediment textures in fig. 3 are grouped according to the classification of Shepard (1954). Silty sand and sandy silt bound the corridor to the northwest and southeast and a tongue of silt from the north extends around Pt. Reyes (Figure 3). More detailed analysis of the sediment texture reveals a slightly more complex regional distribution as shown, for example, by a plot of mean grain size (Figure 4). The increased complexity is well illustrated by examining the cross-shelf corridor of sand defined in Figure 3. Plotting the mean grain size at 1-phi intervals shows that patches of medium and coarse sand exist within a field of fine sand and that sediment texture becomes coarser closer to the Farallon Islands (Figure 4). Increased sampling density reveals an even more complex pattern of sediment texture. Note the area of dense stations (171 spaced 1 km apart) on Figure 1 and Figure 2. Sediment in this area, based solely on analysis of samples from the regional grid of stations, is uniformly fine sand (Figure 4). However, data from the 171 sample grid of stations (also plotted at a 1-phi interval) shows that the area actually consists of a complex pattern of mean grain sizes that range from fine to very coarse sand (Figure 5). This level of sampling density provides data important to interpretation of the modern day depositional and oceanographic processes operating in the Gulf of the Farallones.

REFERENCES

Carver, R. E., 1971, Procedures in Sedimentary Petrography: New York, John Wiley and Sons, 653 p.

Chin, J.L, Rubin, D.M., Karl, H.A., Schwab, W.C., and Twichell, D.C., 1989, Cruise report for the Gulf of the Farallones cruise, F1-89-NC, F2-89-NC off the San Francisco Bay Area, January 6 through 28, 1989: U.S. Geological Survey Open-file Report 89-317, 4. p.

Folk, R. L., 1968, Petrology of Sedimentary Rocks: Austin, University of Texas Publication, 170 p.

Inman, D. L., 1952, Measures for describing the size distribution of sediments: Jour.Sedimentary Petrology, v. 22, #3, p. 125 - 145.

Jordan, F.J., Jr., Fryer, G.E., and Elze, H.H., 1971, Size analysis of silt and clay by hydrophotometer: Jour. Sedimentary Petrology, v. 41, p. 489-496

McHendrie, G., 1988, sdsz - A Program for Sediment Size Analysis: U.S. Geological Survey, Branch of Pacific Marine Geology, Menlo Park, CA.

Shepard, F. P., 1954, Nomenclature based on sand-silt-clay ratios: Jour. Sedimentary Petrology, v. 24, p. 151-158.

Thiede, J., Chriss, T., Clauson, M., and Swift, S.A., 1976, Settling tubes for size analysis of fine and coarse fractions of oceanic sediments: School of Oceanography, Oregon State University, Reference 76-8, 87 p.

Trask, P. D., 1950, Dynamics of sedimentation, in Trask, P. D., ed., Applied Sedimentation: New York, John Wiley and Sons, p. 3-40.