We have compiled an extensive digital geophysical database for the Woodlark Basin region (Fig. F2; Table T1). We have full coverage swath bathymetry and acoustic imagery in the western Woodlark Basin and over much of the eastern Woodlark Basin. More widely spaced ship track data extend onto the margins of the eastern Woodlark Basin and into the Coral and Solomon Seas. Along-track gravity, magnetic, and MCS data are available for many cruises. Land topography and gravity data, as well as teleseismic earthquake locations and focal mechanisms, are also included.

Historic bathymetric surveys are compiled as soundings on Royal Australian Navy and U.S. Defense Mapping Agency charts, many of which we have digitized. Critical hydrographic charts include 19th century British Admiralty surveys around the shoals and reefs bordering the basin, and the 1967-1968 HMAS Moresby survey that charted the region, including Moresby Transform and Moresby Seamount. A seismic sparker and bathymetry survey of Milne Bay was chartered in 1968 (Jongsma, 1972).

Satellite-navigated cruises in the region began with the 1970 Australian Bureau of Mineral Resources, Geology and Geophysics continental margin survey contracted to Compagnie Générale de Géophysique (Willcox, 1973). A reconnaissance survey of ridge subduction in the eastern Woodlark Basin-Solomon Islands region was completed in 1982 on the Kana Keoki (Taylor and Exon, 1987). This was followed in 1985 by a SeaMARC II sidescan and geophysical survey of the triple junction region on the Moana Wave (Crook et al., 1991; Crook and Taylor, 1994).

In 1992, on the Maurice Ewing, scientists completed a MCS, gravity, and Hydrosweep swath bathymetry survey around the D'Entrecasteaux Islands and in the western Woodlark Basin (Mutter et al., 1996). HAWAII-MR1 sidescan and geophysical surveys of the western Solomon Sea and Woodlark Basin were conducted on the Moana Wave in 1992 and April-May 1993, respectively (Galewsky and Silver, 1997; Taylor et al., 1995). These cruises included swath bathymetry and acoustic imagery, gravity, magnetic, and six-channel seismic data along north-south tracks, with a 5-nmi spacing. In August-September 1993, the Mineral Mining Agency of Japan conducted a seafloor mineral survey of the eastern Woodlark Basin using the Hakurei Maru No. 2 (Deep Ocean Resource Development Co., 1995). Hydrosweep swath bathymetry and acoustic imagery and magnetic data were collected along north-south tracks with a spacing of 2.5 nmi. The combined surveys resulted in total bathymetry and acoustic imagery coverage of the entire spreading system, extending from margin to margin west of 154º45'E and farther east at least 50 km on either side of the spreading system.

In November 1995, as part of an ODP site survey, we completed a MCS survey of the region around Moresby Seamount on the Maurice Ewing (Taylor et al., 1996). Hydrosweep swath bathymetry and acoustic imagery and magnetic, and gravity data collected along track provided improved definition of Moresby Seamount. In addition, a transit track crossed the eastern Woodlark Basin. Additional data from the U.S. National Geophysical Data Center (NGDC) archives and satellite gravity and predicted bathymetry values (Sandwell and Smith, 1995; Smith and Sandwell, 1997) complete our suite of marine geophysical data. MCS data were processed using Promax software. Compilation, processing, and display of data utilized the "GMT" software of Wessel and Smith (1995).

A complete bathymetric map of the Woodlark Basin region was compiled using swath bathymetry, wide-beam profiler data, and digitized depth soundings. Where data remained sparse, satellite-predicted bathymetry values were added (Smith and Sandwell, 1997). The data were initially median filtered and placed on a grid of 0.01° latitude and longitude. The higher resolution data, on a grid of 0.002°, was integrated with this data set. The final stage in the data compilation involved the addition of U.S. Geological Survey (USGS) 30"-arc grid topographic data of the islands and the Papuan Peninsula.

Processing of the HAWAII-MR1 acoustic imagery was conducted by the Hawaii Mapping Research Group (HMRG). In addition, we removed erroneous values at the nadir, used an asymmetric (0.1° across the ship track by 0.001° along the ship track) Gaussian filter to extrapolate data into the resulting data gap, and provided corrections to the across-track amplitude variation to improve the continuity of the high backscatter region associated with the spreading center. The final product was a 0.001° grid. The HAWAII-MR1 data were merged with Hakurei Maru No. 2 Hydrosweep acoustic amplitude data from the eastern Woodlark Basin, also on a grid of 0.001°.

Shipboard magnetics data are primarily from the 1993 Moana Wave and Hakurei Maru No. 2 surveys. The data from these cruises had no significant discrepancies between them and were used as a reference. Crossover corrections to other shipboard data were made by sampling the grid of the Moana Wave and Hakurei Maru No. 2 data at the location points of the respective cruises. The average difference between data from each cruise and the grid of the reference data was removed on a cruise-by-cruise basis. The crossover corrected data were then median filtered and placed on a grid of 0.01°. The magnetic anomaly grid was resampled to 0.002° and inverted with the bathymetry grid to derive a magnetization solution assuming a 1-km-thick source layer that conforms to the seafloor, following the technique of Macdonald et al. (1980).

Shipboard free-air gravity anomaly data were crossover adjusted in a similar way to the magnetic data, and discrepancies were removed on a cruise-by-cruise basis with respect to the Maurice Ewing 1995 crossing of the Woodlark Basin. This cruise was used as a reference because of the high quality of its gravity data (collected with a BGM-3 gravity meter; Bell and Watts, 1986) and navigation (using the Global Positioning System), and because it crosses the entire basin from west to east, intersecting the track of most other cruises. First, the data from each cruise that had an intersection with the Maurice Ewing 1995 track were adjusted. Remaining cruises were adjusted by removing the average discrepancy of each of these with respect to the corrected ensemble of the first group. In the case of the Moana Wave 1993 cruise, intermittent problems with the gravimeter created abrupt offsets in the gravity data. Therefore, for this cruise, the data were divided into four segments bounded by the offsets, and each segment was adjusted individually to remove the discrepancy with respect to previously adjusted cruise data.

In addition to shipboard gravity data, we have also merged satellite altimetry-derived gravity data from Sandwell and Smith (1995) to help interpolate the gravity field between tracks. The gridded satellite altimetry-derived gravity data were sampled at the location of the adjusted ship data, and the average difference was removed from the satellite data. Next, a combined data set was compiled by using the adjusted shipboard measurements and gradual tapering to the corrected satellite-derived values over a distance of ~5 nmi. This was done to preserve the higher resolution of the ship data (Neumann et al., 1993) while taking advantage of the uniform coverage of the satellite data to interpolate between widely spaced ship tracks, primarily in the eastern basin.

Using the gridded bathymetry, the gravity effect of the seafloor topography was forward calculated, assuming a density contrast between the water and basement of 1800 kg·m^-3, following Parker's (1973) method (using five terms). Because this method requires the observation plane (sea level) to be above all the relief, the gridded surface was truncated at a water depth of 100 m before the calculation. The gravity effect of the topography was then subtracted from the free-air anomalies to produce Bouguer anomalies. This calculation overestimates the gravity contribution of the topography where there is significant sediment thickness with lower densities than the crustal value we have adopted (2800 kg·m^-3), thereby creating excessively negative anomalies in these areas. The Bouguer gravity anomaly map is particularly useful to examine the differences between the oceanic domains of the Woodlark Basin that have little sediment cover.