The Seafloor Revealed

Methods

Bottom Samples

Between 1984 and 1991, 1,773 bottom sample stations were occupied, resulting in the collection of 1303 bottom samples. Two attempts were made at each station where the sampler initially returned empty, after which the site was considered a rock bottom.

The bottom sampler used was a Smith-McIntyre stainless steel device that nominally collected up to 0.25 m³ of sediment (Figure 13). In mud, the sampler did gather 0.25 m³ of sediment, usually with the surface completely undisturbed (Figure 14). When the sampler was used over a sandy bottom it usually returned an undisturbed sample, unless a large shell blocked its jaws, permitting material to wash out. Over a gravel seafloor it was common for large clasts to prevent closure of the sampler's jaws, resulting in loss of some or all sediment. In those situations, up to two additional attempts were made to obtain a sample before abandoning the station. At stations where no sample could be collected, a hard bottom was inferred and rock was mapped.

Southwest of Cape Small samples were generally collected from grid intersections with a 1 nautical mile (1.85 km) distance between sample sites. Focus was placed on the large sandy embayments off Wells, Saco, and the Kennebec River mouth, as well as in muddy Casco Bay. Relatively few bottom samples were gathered off rocky areas like Kennebunk or Kittery. Geophysical tracklines were later run over the sample stations to permit extrapolation of the bottom-sediment data. North and east of Cape Small, geophysical data were generally gathered before bottom samples. This resulted in a need for fewer samples, and so fewer stations were occupied.

Following collection, samples were stored in coolers until they could be stored in a freezer in the sedimentology laboratory at the University of Maine. Depending on the level of funding or specific needs of a particular project, samples were analyzed for grain size, organic carbon and nitrogen, carbonate content (shells) and heavy mineral concentration (Figure 15). Standard laboratory techniques were employed for the textural analyses, with pipette methods to evaluate the percent of sand, silt, and clay (Folk, 1974), a settling tube to evaluate sand size distribution (based on grain settling velocities), and a Micromeritics Sedigraph to measure the settling velocity of mud-size material (< 62 microns). Carbon and nitrogen were analyzed on a Carlo-Erba Model 1106 Elemental Analyzer, and carbonate content was determined through acid digestion (Molnia, 1974). Heavy mineral content was measured through heavy liquid analysis (Kelley and others, 1987a) or by means of a Humphrey Spiral (Luepke and Grosz, 1986; Lehmann, 1991; Malone, 1997).

Side-Scan Sonar Images

Analog side-scan sonar records along 3,358 km of the seafloor were gathered with an EG-&-G Model 260 slant-range corrected device operating with a Model 272-T towfish at a nominal frequency of 105 kHz (Figure 16a). The device was most often run at a 100 m range (200 m wide swath beneath the research vessel), although ranges from 25 to 300 m were occasionally employed (swaths from 50 to 600 m beneath the boat).

Interpretation of the side-scan sonar records was aided by ground-truth information from the bottom samples as well as from more than 50 submersible dives (Belknap and others, 1988). Although objects as small as lobster traps and current ripples were visible at the 100 m range, it was not possible to make detailed textural distinctions using acoustic imagery alone or to directly compare acoustic images with samples that were analyzed for grain-size distribution. Thus, sandy mud and muddy sand, which are textural categories that can readily be distinguished with particle size analyses (Folk, 1974), are essentially identical in acoustic images. Where sand gradually mixes with mud, a contact was drawn at the midpoint between known occurences of sand and mud. Similarly, where grain-size data were lacking, rippled seabeds were called gravel, even though sand is commonly a minor component of the bedforms.

The heterogeneity of the seabed at all scales precluded mapping all features observed in the side-scan sonar records. To be visible on a map, a feature must be at least 1 mm². This means that on a 1:100,000 scale map, the smallest mappable unit on the seafloor must be at least 10,000 m². Because outcrops of bedrock and gravel smaller than 10,000 m² commonly punctuate generally muddy or sandy areas, it was often necessary to map texturally composite features (Figure 17).

On side-scan sonar images, rock, mud, gravel, and sand usually produce distinct acoustic returns, and so were mapped as distinct units (Figure 17, Figure 18). Rock yields a strong surface return (dark on side-scan sonar records) often with great bathymetric relief and fractures that result in areas with acoustic shadows (Figure 18, Figure 19). Gravel deposits also produce a relatively strong acoustic return (black to dark gray on side-scan sonar records). They are often closely associated with rock, but lack relief and fractures and, instead, are often covered with ripples or boulders (Figure 18, Figure 20). Sand produces a much weaker acoustic return (light to dark gray on side-scan sonar records) than either gravel or rock, and usually lacks local relief (Figure 18). Mud yields a very weak surface return (light gray to white on side-scan sonar records) and, except where it accumulates on steep slopes or near gas-escape pockmarks, it is associated with a smooth seabed (Figure 12).

When sand, gravel, rock, and mud were greater than 10,000 m² in area, they were mapped as separate units. In many places, however, a heterogeneous seabed composed of numerous small features required composite map units (Figure 17). In an area where no single seafloor type exceeded 10,000 m², a composite unit was defined. The dominant end member of sand, mud, rock, or gravel was modified by a subordinate end member (Figure 19, Figure 20). Further subdivision on the relative abundance of one or another bottom types was not possible.

Seismic Reflection Profiles

Seismic reflection profiles were gathered along 5,011 km of tracklines, often in conjunction with side-scan sonar data. A Raytheon RTT 1000a 3.5/7.0 kHz unit with a 200 kHz fathometer trace was used mainly in relatively shallow water over muddy bottoms, while an ORE Geopulse "boomer" seismic system was most effective in deeper water over thicker deposits of sandy or gravelly sediment (Figure 16b).

Nine seismic facies are described from the western Gulf of Maine, seven of which are occasionally exposed at the seabed (Table 1; Barnhardt and others, 1997; Belknap and others, 1989; Belknap and Shipp, 1991). Bedrock (BR) forms the acoustic basement in the area, but is commonly exposed on the seafloor. It is recognized by its intense, sharp initial acoustic return, and high-relief surface (Figure 7, Figure 10, Figure 21). It is frequently overlain by till (T), which also produces a strong surface return. When it is thick, the mound shape of the till or the chaotic internal reflections distinguish it from bedrock; when till is thin, seismic reflection data alone may not always distinguish rock from till (Figure 8).

Glacial-marine muddy sediment (GM) may overlie till or bedrock, and also is commonly exposed at the seafloor (Figure 10). This material provides an intermediate surface return, and ranges from well stratified to acoustically transparent. In depths less than 60 m, it is often unconformably overlain by modern mud (M). Modern mud has a very weak surface return and is typically acoustically transparent (Figure 7, Figure 11).

Deltaic (D) and estuarine (E) sediments from the late Pleistocene to early Holocene occur near some large river mouths (Barnhardt and others, 1997; Figure 10, Figure 11). These materials produce strong surface returns, and usually have good internal stratification. They are usually covered by a reworked sand and gravel deposit (SG), or modern mud (M). Thin gravel layers (TGL) and natural gas deposits (NG) are also recognized beneath deposits of sand and mud, respectively. These acoustic units are never exposed at the seafloor, although natural gas has erupted from the seabed in some locations (Figure 12; Kelley and others, 1994).

Although seismic reflection profiles are most useful in constructing the geologic history of an area, the bathymetry and geologic context provided by the seismic reflection profiles, along with the strength of the surface return, also allows identification of the surficial deposit. When used in conjunction with the side-scan sonar, both the age and nature of the surficial sediment are easily interpreted (Figure 21).

Navigation and Compilation

Navigation fixes in the outer estuaries and offshore areas were made every 2-5 minutes with LORAN-C. LORAN coordinates were later converted to latitude/longitude with the computer program, LORCON, and provided an accuracy of ± 100 m (J. Stuart, NOAA, personal communication). In the upper reaches of the estuaries, navigation fixes were established with line-of-site observations, radar measurements, and visual observations of buoys and landmarks. The accuracy based on these observations varied from less than ± 10 m to around ± 200 m. Recent work in Cobscook Bay, Penobscot Bay, Wells Embayment, and in the Kennebec River utilized a Global Positioning System (GPS) and was accurate to ± 10 m.

All navigation was converted to Universal Transverse Mercator (UTM) projection and plotted through a geographic information system (GIS), ARC/INFO (UNIX version 7.03). The shoreline of the region and bathymetry (10 m contour interval) were digitized from NOAA Bathymetric Charts (Table 2). The charts are only provisional blue-line paper copies for most of the region, but they provide a 2 m contour interval in many locations. Difficulty in interpretation of positive and negative changes in bathymetry on the poorly labeled charts created many possible errors, especially in areas where we lacked accompanying geophysical data. Partly for this reason we caution users of the maps that they are not suited for navigation purposes.

The surficial geologic maps were prepared by overlaying the side-scan sonar navigation fixes on the bathymetry in the GIS. A buffer, or area equal to the observational range (swath) of the side-scan sonar instrument, was drawn parallel to the navigation fixes and plotted. Surficial geology was interpreted from the original side-scan sonar records and transferred by hand onto a mylar cover sheet that was itself later digitized into the GIS. Where the spacing of the side-scan sonar lines was more than twice the width of the range (i.e. no overlaps), the surficial geology between the lines was interpolated with the aid of the bathymetry, bottom samples, and seismic reflection profiles (where they existed). The directly interpreted swath areas have been mapped in bright colors (Barnhardt and others, 1996a-g). Each color represents one of the primary map units (mud, gravel, sand, rock). Patterns over the colors relate to the specific composite map unit (Figure 17) that was observed.

Where side-scan sonar data were scarce or absent, reliance was placed on seismic reflection records and bottom samples in conjunction with bathymetry. These data are depicted in dull colors on the maps and lack patterns since detailed textural information could only be discerned from the side-scan sonar data (Barnhardt and others, 1996a-g, 1998). A physiographic map was prepared largely on the basis of bathymetry with supplementary information provided from the geophysical data (Kelley and others, 1989a; Kelley and Belknap, 1991; Barnhardt and Kelley, 1995; Kelley and others, 1997). Feature maps (an inset to the surficial geologic map series) were compiled from NOAA charts for buoys, cable crossings, disposal sites, and and limits of the territorial sea (Table 2). Shipwreck coordinates were taken from the National Ocean Survey Automated Wreck and Obstruction Information System (Table 3, NOAA, 1990, 1998). The accuracy of most wreck positions is unknown. Gas fields and gas-escape pockmarks were interpreted from side-scan sonar and seismic reflection profiles (Kelley and others, 1989c, 1994).

Contents Introduction Methods Physiography Surficial Geology Geological History Summary Glossary References

Last updated on October 6, 2005