4.2. The 210Pb Fluxes
 Inventories of excess 210Pb vary markedly across the GOP, although they generally decrease with increasing water depth (Figure 5). This spatial variation can be understood by comparing the sources of 210Pb in the region with the fluxes of 210Pb required to support excess 210Pb inventories at steady state conditions (equation (4)).
 The primary sources of 210Pb to waters above the shelf break and slope are atmospheric fallout of 210Pb to surface water, and production from decay of 226Ra [Cochran, 1982; Cochran et al., 1990]. By comparison, inputs of dissolved 210Pb from rivers are small, mainly because of rapid scavenging of dissolved Pb by suspended sediments [Benninger, 1978]. Assuming efficient trapping of river-borne sediments on the inner shelf, the total expected inventory of excess 210Pb deposited on the distal seabed is the sum of the two primary sources mentioned above. We refer to the sum of flux from these two sources as the theoretical flux (T).
 The mean annual atmospheric flux of 210Pb, measured in rainfall at Townsville, northeastern Australia, between 1964 and 1970, was 0.22 dpm cm−2 a−1 [Bonnyman and Molina-Ramos, 1975]. This average atmospheric 210Pb flux is in good agreement with the 210Pb flux of 0.23 ± 0.07 dpm cm−2 a−1 reported by Turekian et al.  from the same area. However, J. Pfitzner (unpublished data, 2001) measured 210Pb fluxes in rain collections at the Australian Institute of Marine Sciences (Townsville) and reported a 210Pb flux of 0.33 dpm cm−2 a−1. We note that temporal and spatial variations in rainfall cause fluctuations in the atmospheric 210Pb flux, and we use a value of 0.3 dpm cm−2 a−1 [Brunskill et al., 2003; Pfitzner et al., 2004] as representative atmospheric 210Pb flux for our study area.
 We did not measure 226Ra activities in the water column from the study area. Consequently, we estimate water column inventory of 226Ra based on measurements from GEOSECS Station 263 located at 16°40′S; 167°05′W in the southwestern Pacific, integrated over the depth range at each core location. The profile of 226Ra at this station increases to a maximum above 1500 m water depth, decreases below 2000 m, and activity of 226Ra ranges between 6.0 ± 0.7 and 27.3 ± 0.6 dpm 100 kg−1 from surface to 2500 m water depth [Chung and Craig, 1980]. Water column generation of 210Pb from 226Ra is calculated as the product of 226Ra water column inventory and the 210Pb decay constant (assuming negligible loss of the intermediate decay product 222Rn).
 For the present discussion, we will define the ratio between the core-calculated flux (F) (equation (4)) and theoretical flux (T) as R = F/T. If R = 1, then observed inventories are consistent with vertical scavenging of 210Pb from the water column inventory, whereas values of R > 1 suggest focusing or lateral import of 210Pb to the area. Similarly, a value of R < 1 indicates lateral export of 210Pb from that area [e.g., Buesseler et al., 1985; also see Baskaran and Santschi, 2002].
 The highest values of R occur along the upper slope and shelf edge of the northeastern Pandora Trough, where R ≥ 10 for six cores in water depths of 90–1100 m, and R > 5 for three additional cores in the area (Figure 2 and Table 1). In the Ashmore Trough, intermediate values of 2 ≤ R ≤ 5 suggest modest enrichment of 210Pb flux with respect to vertical scavenging of in situ production. These cores (MV-10, MV-12, MV-14, and MV-16) were collected seaward of escarpments of Portlock, Boot, and Ashmore reefs. Cores from the Moresby Trough, southwestern Pandora Trough, and the Eastern Plateau display the lowest values of R (0.9 ≤ R ≤ 2), suggesting that in these areas, approximate equilibrium exists between in situ water column production, and seabed inventories. These cores are from the deepest basins, and are the most remote from the shelf edge and escarpments, in our study (Table 1 and Figure 1).
 One potential source of additional 210Pb is the northern extension of the Coral Sea Current [Wolanski et al., 1995], which could provide 210Pb through the process of “boundary scavenging” [Anderson et al., 1994; Moore et al., 1996; Baskaran and Santschi, 2002]. This tropical oceanic water mass is likely to be very low in suspended particulates that could scavenge dissolved 210Pb for export to the seabed. Because GOP is a semienclosed basin, circulation is strongly influenced by the bathymetry and the coastal geometry, with substantial transport across depth contours [Wolanski et al., 1995; Ogston et al., 2008; Slingerland et al., 2008b]. More abundant biogenic and lithogenic particles found closer to the GOP shelf could thus scavenge this dissolved 210Pb advected from offshore waters, and then the particles could settle to the seabed. Boundary scavenging of 210Pb from such horizontally advected water masses might be further enhanced by along-slope barriers to flow, such as the escarpments surrounding Ashmore Trough (Figure 2), and, in the northeastern Pandora Trough, the sharp bend in the shelf edge and pronounced steepening of the slope to the southeast (Figures 1, 2, and 7), the two regions where the greatest relative enrichments of 210Pb flux are observed. G. J. Brunskill and J. Pfitzner (unpublished data, 2006) measured an inventory of excess 210Pb from sediment traps at the base of the continental slope that was 10–20 times the expected atmospheric input plus in situ decay supply rates, consistent with our core observations and calculations in Table 1.
 An additional potential source of 210Pb to seabed inventories is sediment focusing, as observed in mud deposits off some rivers [e.g., Bruland et al., 1974; Krishnaswami et al., 1975; Carpenter et al., 1981; Buesseler et al., 1985; Thorbjarnarson et al., 1986; Smoak et al., 1996; Kuehl et al., 2004]. SAR and 210Pb inventories tend to display positive correlations in these settings, producing sediment depocenters with 210Pb inventories elevated above values expected from the theoretical vertical flux. Sediment focusing is probably a factor in the northeast Pandora Trough (where SARs are relatively high), more so than in the Ashmore Trough (where SARs are very low). If in situ decay of 226Ra in the water column is a dominant source of 210Pb, inventories would be expected to increase progressively offshore, assuming accumulation rates were uniform. Therefore, from the Pandora Trough into the deeper Moresby Trough, the observed relative downslope decrease in estimated 210Pb fluxes is interpreted to reflect a similar decrease in apparent sediment accumulation rates, resulting from sediment focusing and retention along the shelf edge.
 On the basis of recent studies, bathymetric data, and seismic profiles of the shelf and troughs [e.g., Febo et al., 2008; Francis et al., 2008; Patterson et al., 2006; this study], the primary source of siliciclastic material for both Pandora and Moresby troughs appears to originate from the northeastern corner of the GOP. For example, calcium carbonate content in multicore tops in the GOP illustrate that the lowest carbonate contents (<45 wt%) are observed on the shelf edge and upper slope of the northeast corner of Pandora Trough (pelagic production of carbonate), and along a channel in Moresby Trough (Figure 6) [Febo et al., 2006]. In contrast, calcium carbonate exceeds 55–60% in the southwestern Pandora Trough and amounts to as much as 90% in the northern Ashmore Trough (sources of both neritic and pelagic carbonate) (Figure 6). This pattern is generally consistent with sediment focusing in the northeastern Pandora Trough and decreasing siliciclastic particle flux toward the carbonate-dominated Ashmore Trough (Figures 6 and 8). The most likely source of this sediment is the aforementioned extensive subaqueous clinoform of the GOP inner shelf (Figure 1), produced by fluvial sediment discharged from the Fly, Kikori, and Purari rivers, among others [Walsh et al., 2004]. The shelf edge depocenter, described above near locations of cores MV-42 and MV-47 is probably fed by oceanic transport of inner shelf sediment from the northwest. Sediment accumulation near the shelf edge and steep upper slope may lead to episodic production of turbidity flows that move downslope, toward the channel in the Moresby Trough, as recorded in cores MV-24 (Figure 2) and MV-22 [Patterson et al., 2006].
Figure 8. Map of relative enrichment in excess 210Pb inventories from equation (4) and Table 1, showing submarine channel network derived from DEM analysis (channel network data from Patterson et al.  with permission). The largest dots indicate highest relative levels of 210Pb enrichment, primarily on the shelf edge and upper slope of the northeast GOP. Patterns of 210Pb enrichment and channel networks outline important sediment transport pathways from the shelf edge to deep basins.
Download figure to PowerPoint
 A network of submarine channels in the Pandora and Moresby troughs, capable of delivering gravity-driven flows from the shelf edge to Moresby Canyon, is shown in Figure 8, along with values of R for individual cores (Table 1). The channel network was extracted from the bathymetric grid of Daniell  using RiverTools© software (RIVIX LLC) for DEM analysis. The resulting channel network identifies potential pathways for sediment transport that link the locations of cores MV-42, MV-47, and MV-24, among others, and gives a representation of possible flow pathways in shelf and slope settings where particles may move into the basin by gravity flow transport [Bird et al., 1995].
4.3. A Sediment Transport Scenario
 Sediment focusing and boundary scavenging of oceanic 210Pb are both likely sources for producing elevated 210Pb inventories on the shelf break and upper slope. Both processes are probably active in the northeastern Pandora Trough, but boundary scavenging is probably dominant to the southwest, in regions of lower suspended-sediment flux. Spatial distribution of 210Pb flux observed at the seabed can be attributed to downslope and lateral particle transport and variation in bottom topography. This phenomenon has been observed in a number of continental slope environments, e.g.: Quinault Canyon [Carpenter et al., 1981; Thorbjarnarson et al., 1986], California margin [Huh et al., 1990; Alexander and Venherm, 2003] and the eastern and southeastern U.S. continental margin [Biscaye et al., 1988; DeMaster et al., 1991]. A certain proportion of particles may be transported from shelf to open slope waters and then down slope, and the rest may be carried by along-shelf currents to canyon heads, settling into canyons, thus facilitating down canyon transport [e.g., Biscaye and Anderson, 1994].
 The northeastern GOP shelf edge is marked by several locations that could be major escape routes for sediments from the shelf to deeper basins (Figure 8). For example, cores MV-18 and MV-19 are situated at the head of an intraslope canyon as shown by analysis of seafloor bathymetries and channel networks (Figures 2 and 8). Further, the highest measured excess 210Pb inventory from this study is located at MV-18, along with low surficial calcium carbonate content (Figures 6 and 8). Cores MV-42 and MV-47 also occur along a channel extending from shelf to basin, and both cores exhibit enriched 210Pb inventories and relatively high SAR. Farther down the same channel network, core MV-50 (795 m water depth) exhibits the same characteristics, suggestive of sediment focusing and downslope transport. Combined seafloor bathymetry and radionuclide data thus suggest that these locations are major conduits for sediment supply to deeper basins. These observations, combined with evidence from core MV-24, further suggest that sediment transport occurs in both nepheloid layers and turbidity flows.