propose appropriate stratigraphic models (e.g., Andrews and Matsch, 1983; Gilbert, 1983; Molnia, 1983; Elverhoi, 1984; Powell, 1984; Eyles et al., 1985; Dowdeswell, 1987; Powell and Molnia, 1989), yet there is very little information, on a regional scale, of the flux of sediment within the glacial marine environment, especially on time scales of a few thousand years. However, Ruddiman (1977) and Fillon (1985), amongst others, computed the net flux of sand-size particles (ice rafted detritus, IRD) to the North Atlantic deep basins. We confine our discussion, therefore, to areas where we have worked and where such data are available.
This chapter presents data on the flux of sediment to the seafloor along fiord to shelf transects. It thus forms a link between current debates on the efficiency and magnitude of glacial erosion and transfers of sediment to deep sea basins (cf., Laine, 1980; Bell and Laine, 1985; Andrews et al., 1985a). The chapter has two approaches: in the first section we examine the processes that produce sediment along glaciated continental margins and suggest a space-time ranking of these for sediment accumulation. We also examine the climatic and glaciological controls on sediment flux. In the second part, we illustrate some of these aspects by examining two specific regions, the first being the borderland of the eastern Canadian Arctic, fronting Baffin Bay, and the second being part of the East Greenland shelf. The latter area is an analog of the conditions that prevailed in the Canadian Arctic during the last glacial cycle. These areas represent subpolar to polar glaciological environments and contrast (see later) with more active environments in southern Alaska.
Sediment flux is defined as sediment mass per unit area with respect to time. We normally measure the net flux of sediment to the seafloor, that is kg/m2 per unit time, hereafter termed Fn↓. Consider Figure 7.2 — this shows the direction of sediment fluxes along a fiord to shelf transect. The vertical flux of sediment between two points along this transect is defined as the level of reduced suspended sediment inventory (I in units kg/m2) throughout the water column between these two points. If dI/dt = λI, where λ is a first order removal rate constant, then Ff↓, defined as the net downward flux of sediment can be expressed as Ff↓(x) = Ff0e-λt, where Ff0 is maximum sedimentation rate at the fiord head (Syvitski et al., 1988).
However, we have an additional flux of sediment (Fs↓) being contributed from the shelf (iceberg melting, current resuspension, iceberg ploughing, etc.) and also from local disturbances (F1↓), say between x3 and x4 (Figure 7.2), due to gravity flow deposition which delivers sediment to the fiord basins (Syvitski, 1989). Thus, the net sediment accumulation
is the resultant of the three sources plus allochthonous (carbon and silica) deposition, minus losses due to erosion (Fr), so that:
and over some interval of time the total mass of sediment (kg per meter width per unit time) contributed to the sea-floor would be:
In this chapter we ignore the carbon flux to the sea floor because, although it is important, it is usually less than 1 percent dry weight of the sediment (Andrews, 1987a; Syvitski et al., 1989, 1990). A potentially more serious matter is the absence of a measure of the biogenic silica flux, although we know from the work of Williams (1988) that this can be substantial.
Two critical factors become important in determining the amount of suspended sediment that escapes from the fiord to the adjacent shelf: (1) the length of the fluvial plume, and (2) the length of the fiord basin. The plume length is primarily a function of the freshwater discharge but the fiord length is an independent variable. For example, the fiords of southwest Greenland are about twice as long as the northeast Baffin fiords, fluvial input is 2 to