Fig. 1. General bathymetry of the Ross Sea region [60]. Contour interval is 50 m between 0–1000 m and 500 m between 1000–5000 m. Geological features discussed in the text are located, along with the position of the Cape Roberts and DSDP drilling sites. The thin, white lines locate the main sedimentary basins and basement highs comprising the Ross Sea. Apart from the bathymetric escarpment separating the transition from the Ross Sea to the Pacific Ocean (and the approximate location of the ocean–continent boundary), the most conspicuous features of the bathymetry are northeast–southwest glacial scars.

Fig. 2. Bouguer gravity anomaly of the Ross sea region [60]. Reduction density is 2400 kg/m³. Contour interval is 10 mGal. The location of the main sedimentary basins (thin white lines) and interbasinal highs are superposed and clearly show a general anti-correlation between Bouguer gravity high and the location of sedimentary basins. Basement highs are associated with gravity lows. Location of the late Eocene–Neogene Terror Rift is shown by the bold yellow line while the surface outcrop of alkali basalts is shown by the thin red line. Seismic profiles shown in Fig. 3 and Fig. 4 are also located.

Fig. 3. Seismic reflection profile and corresponding Bouguer gravity anomaly profile across the Victoria Land Basin, Central Trough, and Eastern Basin of the Ross Sea [24]. Clearly shown is the anti-correlation between basin depocenters and positive Bouguer gravity anomalies.

Fig. 4. Depth to magnetic basement (gray circles) superimposed on basement (bold black line) mapped from seismic reflection data across the western Ross Sea. The Bouguer gravity [60] and magnetic data from the GANOVEX IV mission [42] are shown for reference. The gray circles are the Werner deconvolution solutions for depth to magnetic basement with the size of the circles being scaled to magnetic susceptibility.

Fig. 5. Kinematic and isostatic modeling procedure for extensional basins and passive margins. (a) Crustal stretching factor, β(x), for single and multiple faulting. After faulting, the resultant crustal thickness is t_c(1 − 1 / β(x)). The form of β(x) mirrors the distribution and amplitude of accommodation generated across the region. The basin-bounding fault of the system is at 800 km and after 1200 km, the crustal extension is a constant β(x) of 2. (b) Kinematic description of simple slip along a single border fault that produces a basin by the collapse of the hanging wall block. The geometry of the fault governs the size and shape of the resulting basin. (c) Kinematic description of multiple faulting of the crust. (d) Local isostatic adjustment of the crustal extension shown in (c) assuming water as the infill material and generation of the compensating Moho topography. (e) Local isostatic adjustment of the crustal extension shown in (c) assuming sediment as the infill material and generation of the compensating Moho topography.

Fig. 6. Modeled gravity across a series of rift depocenters simulating the general geometry and distribution of basins comprising the Ross Sea region (scheme shown in Fig. 5d). Water, sediment, crustal and mantle densities are assumed to be 1030, 2400, 2800 and 3300 kg/m³, respectively. (a) Basin geometry and modeled gravity assuming local isostasy and no sedimentation. Major depocenters are characterized by a negative gravity effect. (b) Basin geometry and modeled gravity assuming local isostasy and sediment infilling. The resulting deflection locally depresses and subdues the Moho topography. Again, major depocenters are characterized by a negative gravity effect. (c) Basin geometry and modeled gravity assuming that the infilling sediments load a crust with finite flexural strength (T_e of 30 km). In this case, the deflection regionally depresses the Moho topography. In particular, the Moho relief is preserved rather than attenuated as in the low-strength case. Now, the major depocenters are associated with a gravity high while basement highs are characterized by gravity lows.

Fig. 7. Modeled free-air and Bouguer gravity across a series of rift depocenters simulating the general geometry and distribution of basins comprising the Ross Sea region (scheme shown in Fig. 5d). Water, sediment, crustal and mantle densities are assumed to be 1030, 2400, 2800 and 3300 kg/m³, respectively. (a) Basin geometry and modeled free-air (bold black line) and Bouguer (thin red line) gravity assuming a rift effective elastic thickness, T_e, of 5 km and no sedimentation. Major depocenters are characterized by a negative free-air gravity. In contrast, the Bouguer gravity is dominated by the local and regional variations of the Moho. (b) Basin geometry and modeled free-air and Bouguer gravity assuming an effective elastic thickness of 5 km during both rifting and sedimentation. Major depocenters are characterized by negative free-air and Bouguer gravity. (c) Basin geometry and modeled free-air and Bouguer gravity assuming that the infilling sediments load a crust with a T_e of 30 km while the rifting event is associated with a T_e of 5 km. The major depocenters are characterized by gravity highs (both free-air and Bouguer) while the basement highs are characterized by gravity lows.

Fig. 8. Process-oriented gravity modeling investigating the contribution of magmatic underplating and crustal intrusions to the gravity anomaly. Local isostasy is assumed during crustal thinning. During sedimentation, magmatic underplating and crustal intrusions the effective elastic thickness is assumed to be either zero (local isostasy; left-hand column) or a T_e of 30 km (right-hand column). (a) Gravity effect across a modeled sedimentary basin assuming local isostasy during extension and either local isostasy (left-hand column) or a T_e of 30 km (right-hand column) during sedimentation. For locally compensated systems, the gravity negative correlates with the localization of the basin depocenter. In contrast, for sedimentation with T_e of 30 km, the gravity across the basin depocenter is positive. (b) Gravity effect across a modeled sedimentary basin assuming local isostasy during extension and either local isostasy (left-hand column) or a T_e of 30 km (right-hand column) during magmatic underplating. Underplating induces uplift and for local isostasy, inverts the basin center while adding to the crustal thickness. The resultant gravity is positive but correlates with a zone of major basement uplift within the basin. In contrast, for magmatic underplating when T_e is relatively large, the basin is regionally uplifted and the gravity over the basin depocenter is negative. (c) Gravity effect across a modeled sedimentary basin assuming local isostasy during extension and either local isostasy (left-hand column) or a T_e of 30 km (right-hand column). Crustal intrusions effectively increase the density of the lower crust and induce subsidence. Two intrusion densities were assumed, 2900 and 3000 kg/m³ (red and black lines, respectively). For local isostasy, basin subsidence is amplified and the gravity effect is negative. In contrast, if T_e is significantly large during the intrusion process, the basin is forced to regionally subside. The gravity effect is positive over the basin depocenter.

Fig. 9. Admittance function Z(k) between the Moho and basin gravity as a function of varying rift and sedimentation flexural rigidities. Two families of admittance functions are shown depending on the effective elastic thickness of the extended lithosphere; T_e^r of 5 km (black curves), T_e^r of 30 km (red curves). For T_e^r of 5 km, sediment loading effective elastic thickness, T_e^s, ranges from 5–30 km while for a T_e^r of 30 km, sediment loading effective elastic thickness ranges from 30–80 km. For Z(k) < 1, the gravity effect over the basin is dominated by the density variation of the basin. For Z(k) > 1, the gravity is dominated by the Moho and the basin is associated with a positive gravity effect. When T_e^r and T_e^s are equal, the gravity is always negative across the basin (Z(k) < 1). As the difference between rift and sediment loading effective elastic thickness increases, the gravity over the basin becomes progressively more positive but only for a limited range of wavelengths. The wide late Cretaceous basins of the Ross Sea should be characterized by a positive gravity anomaly while the narrow late Paleogene–Neogene Terror Rift should be associated with a negative gravity anomaly.

Ross Sea basins modeling parameters