Michael A. Summerfield

Modeling postbreakup landscape development and denudational history across the southeast African (Drakensberg Escarpment) margin

[1]xa0We employ a numerical surface processes model to study the controls on postbreakup landscape development and denudational history of the southeast African margin. Apatite fission track data, presented in the companion paper, suggest that the Drakensberg Escarpment formed by rapid postbreakup river incision seaward of a preexisting drainage divide, located close to its present position, and subsequently retreated at rates of only ∼100 m m.y.−1. Numerical modeling results support such a scenario and show that the prebreakup topography of the margin has exerted a fundamental control on subsequent margin evolution. The rheology of the lithosphere, lithological variations in the eroding upper crust, and inland base level falls provided secondary controls. A relatively low flexural rigidity of the lithosphere (Te ≈ 10 km) is required to explain the observed pattern of denudation as well as the observed geological structure of the southeast African margin. Lithological variations have contributed to the formation of flat-topped ridges buttressing the main escarpment, as well as major fluvial knickpoints. Both these features have previously been interpreted as supporting significant Cenozoic uplift of the margin. An inland base level fall, possibly related to back-cutting of the Orange River drainage system and occurring 40–50 m.y. after breakup, explains the observed denudation inland of the escarpment as well as the development of inland drainage parallel to the escarpment. Our model results suggest that in contrast to widely accepted inferences from classical geomorphic studies, the southeast African margin has remained tectonically stable since breakup and escarpment retreat has been minimal (<25 km).

Tectonic uplift, threshold hillslopes, and denudation rates in a developing mountain range

Studies across a broad range of drainage basins have established a positive correlation between mean slope gradient and denudation rates. It has been suggested, however, that this relationship breaks down for catchments where slopes are at their threshold angle of stability because, in such cases, denudation is controlled by the rate of tectonic uplift through the rate of channel incision and frequency of slope failure. This mechanism is evaluated for the San Bernardino Mountains, California, a nascent range that incorporates both threshold hillslopes and remnants of pre-uplift topography. Concentrations of in situ–produced cosmogenic 10 Be in alluvial sediments are used to quantify catchment-wide denudation rates and show a broadly linear relationship with mean slope gradient up to ∼30°: above this value denudation rates vary substantially for similar mean slope gradients. We propose that this decoupling in the slope gradient–denudation rate relationship marks the emergence of threshold topography and coincides with the transition from transport-limited to detachment-limited denudation. The survival in the San Bernardino Mountains of surfaces formed prior to uplift provides information on the topographic evolution of the range, in particular the transition from slope-gradient–dependent rates of denudation to a regime where denudation rates are controlled by rates of tectonic uplift. This type of transition may represent a general model for the denudational response to orogenic uplift and topographic evolution during the early stages of mountain building.

Quantifying passive margin denudation and landscape development using a combined fission-track thermochronology and cosmogenic isotope analysis approach

Abstract A new approach to empirically constraining numerical models of denudation and landscape development across passive continental margins is presented involving a combination of apatite fission-track thermochronology and cosmogenic isotope analysis. This integrated strategy is applied to the Namibian sector of the south-west African margin where the conventional landscape evolution model of inland retreat of an escarpment initiated at the coast at break-up implies a mean rate of retreat of ∼1 km Ma −1 . The thermochronological and cosmogenic isotope data are incompatible with a simple escarpment retreat model since they show that denudation rates oceanward of the present escarpment position have been low ( −1 ) since the end of the Eocene, and the estimated rate of escarpment retreat has been only ∼10 m Ma −1 . This low rate of retreat is, however, consistent with numerical landscape evolution models where the escarpment is pinned at an inland drainage divide.

Denudational history along a transect across the Drakensberg Escarpment of southern Africa derived from apatite fission track thermochronology

[1]xa0The denudational history of a ∼500 km long transect across the Drakensberg Escarpment on the high-elevation passive margin of SE Africa is quantified on the basis of thermal history modeling of apatite fission track data for 15 deep borehole samples, supplemented by an additional 10 outcrop samples. A minimum of 4.5 km of denudation since formation of the margin ∼130 Myr ago is estimated for the coastal zone, with a marked Early Cretaceous episode of accelerated denudation broadly coincident with continental breakup. Samples from the Swartberg borehole (SW 1/67) located ∼30 km seaward of the present position of the Drakensberg Escarpment indicate a total depth of denudation of 3.1 ± 1.2 km since ∼91 Myr, with a phase of accelerated denudation of 2.1 ± 0.9 km at a mean rate of 95 ± 43 m/Myr between ∼91 and 69 Myr. Samples from the Ladybrand borehole (LA 1/68) west of the Lesotho Highlands indicate 1.7 ± 0.5 km of denudation since ∼78 Myr, with a phase of accelerated denudation at 82 ± 43 m/Myr from ∼78 to 64 Myr. Average denudation rates declined to about 10 m/Myr during much of the Tertiary. Although the apatite fission track data do not provide any direct constraints on the denudational history of the Lesotho Highlands, interpolation between the two boreholes, constrained by geological evidence and extrapolated in situ-produced cosmogenic 36Cl-derived denudation rate estimates, suggests a pattern of denudation compatible with numerical modeling studies of escarpment evolution involving rapid river incision seaward of a preexisting inland drainage divide. These patterns of denudation are incompatible with constant retreat of the Drakensberg Escarpment from an initial position near the present coast. We suggest that the Drakensberg Escarpment formed by rapid post-breakup river incision seaward of a preexisting drainage divide located just east of the present escarpment location and became pinned at this divide with subsequent retreat rates of only 100–200 m/Myr.

Geomorphological applications of cosmogenic isotope analysis

Cosmogenic isotope analysis involves the measurement of cosmogenic nuclides that have accumulated in the upper few metres of the Earth’s surface as a result of interactions between cosmic rays and target elements. The concentrations of these cosmogenic nuclides can provide quantitative estimates of the timing and rate of geomorphic processes. In dating applications the concentration of cosmogenic nuclides is interpreted as reflecting the time elapsed since a surface exposure event. However, over most of the Earth’s surface for most of the time the landsurface experiences incremental denudation and in these circumstances cosmogenic nuclide concentrations are related to the rate of denudation. Applications of event dating using cosmogenic isotopes include constructional landforms such as volcanic and depositional features, fault displacement, meteorite impacts, rapid mass movement, bedrock surfaces rapidly eroded by fluvial or wave action or exposed by glacial retreat, and the burial of sediment or ice. Strategies for quantifying rates of incremental change include estimates of denudation rates from site-specific samples and from fluvial sediment samples reflecting catchment-wide rates, and measurements of cosmogenic nuclide concentrations in soils and regolith to quantify rates of rock weathering. The past decade has seen a rapid growth in applications of cosmogenic isotope analysis to a wide range of geomorphological problems, and the technique is now playing a major role in dating and quantifying rates of landscape change over timescales of several thousands to several millions of years.

Long-term rates of denudation in t he Dry Valleys, Transantarctic Mountains, southern Victoria Land, Antarctica bases on in-situ-produced cosmogenic 21Ne.

Abstract Concentrations of cosmogenic 21 Ne ( 21 Ne c ) measured in quartz have been used to estimate long-term rates of denudation for contrasting landscape components in the Dry Valleys area of the Transantarctic Mountains, southern Victoria Land, Antarctica. Samples of Beacon Supergroup sandstones and granitic basement were collected from two contrasting landscape elements—low-relief, high-elevation surfaces and rectilinear slopes—to assess variations in rates of denudation with topographic position. The sample sites for rectilinear slopes were selected because of proximity to 40 Ar / 39 Ar-dated lavas and ash-avalanche deposits. All 21 Ne / 20 Ne ratios are significantly greater than the atmospheric value, and concentrations of 21 Ne c were calculated from the measured 21 Ne values assuming an atmospheric composition for the trapped component. Apparent exposure ages calculated from the concentrations of 21 Ne c , assuming no denudation since exposure, range from 3.78–4.66 Ma for samples from the high-elevation plateau surface, and 0.61–2.48 Ma for samples from the rectilinear slopes. Exposure ages for 3 He c were 2 to 42 times lower than those derived from the abundances of 21 Ne c because of preferential diffusive loss of 3 He c from the quartz lattice; concentrations of 3 He c were, therefore, not used in the calculation of rates of denudation. We interpret the abundances of 21 Ne c as reflecting variations in the rates of denudation rather than exposure age in view of independent evidence for prolonged exposure (>15 Ma) of bedrock surfaces at the sample sites. Calculated maximum rates of denudation range from 0.26–1.02 m Ma−1 for the rectilinear slopes, down to only 0.133–0.164 m Ma−1 for the high-elevation surface sites. These rates are comparable to other estimates of denudation rates for the Dry Valleys derived from analyses of cosmogenic isotopes, but are around two orders of magnitude lower than the long-term mean rate over the past ∼50 Ma estimated from fission-track thermochronology. Combined with the preservation of volcanic deposits dating back to the mid-Miocene, these 21 Ne c data demonstrate that only minimal modification of the landscape has occurred in the Dry Valleys over at least the past ∼15 Ma. This significant conclusion supports the view that the East Antarctic Ice Sheet has been essentially stable over this period rather than experiencing major fluctuations as late as the Pliocene, as has previously been suggested.

Isopach and borehole data as indicators of rifted margin evolution in southwestern Africa

Abstract Isopach and borehole data are used to document post-rifting sediment accumulation along the western margin of southern Africa. The isopach data reveal an increasing rate of sediment accumulation up to the end of the Eocene, but the borehole data generally show the opposite trend. An integrated analysis of both types of data, however, is consistent with a general model of post-rifting drainage development in which the area supplying sediment expanded significantly after breaching of a linear upwarp along the margin allowed a dominant drainage network to develop. The borehole data suggest that such a major sediment outlet from the interior of the sub-continent (the palaeo-Orange) became established in the Late Cretaceous (Coniacian) in the vicinity of the present Orange River mouth. After conversion to equivalent rock volumes, the isopach data can be combined with the general drainage model to infer the likely spatial and temporal patterns of erosion over southern Africa during the accumulation of four offshore sedimentary sequences, beginning with the onset of rift-related sedimentation and extending up to the present. This analysis, using three variant forms of the general model to encompass a wide range of drainage development possibilities, indicates that the average rate of erosion for catchment areas draining to the western margin has declined from 41 – 82 m Ma −1 during the first sedimentary sequence (∼152 to ∼113 Ma) to about 7–9 m Ma −1 during the last sequence (∼37 to 0 Ma). Maximum depths of post-rifting erosion of several kilometres appear to have occurred in a narrow zone bordering the margin, primarily during its early development, while depths of erosion in some interior areas have probably been an order of magnitude less. The average depth of post-rifting erosion for the present-day catchment as a whole is estimated to be about 1.8 km, and the erosion rate over this period has averaged about 12 m Ma −1 .

Landscape development in the Royal Society Range, southern Victoria Land, Antarctica : stability since the mid-Miocene

Abstract Post-rifting landscape development in the Royal Society Range, a rift-flank block in the southern Victoria Land sector of the Transantarctic Mountains, has been reconstructed through a combination of morphological mapping and geochronological data. Creation of the Royal Society Range rift flank ∼55 Ma BP was associated with extension in the Ross Sea Basin and some surface uplift of the Royal Society Range probably occurred at this time. Extrapolation of fission-track data for other sectors of the Transantarctic Mountains, coupled with a reconstruction of pre-rift stratigraphy, indicates that a seaward-thickening wedge of crustal section up to ∼6 km at the coast has been removed since rifting. Much of this crustal stripping probably occurred in the early Cenozoic, and cosmogenic isotope data together with 40 Ar / 39 Ar -dated volcanic cones and surficial ashes demonstrate that denudation over much of the Royal Society Range has been insignificant since the mid-Miocene. This denudation probably occurred primarily through fluvial processes, and the generally limited impact of subsequent glacial action has led to the preservation of elements of the pre-glacial fluvial landscape. The present elevation of a sub-aerially erupted lava flow constrains maximum surface uplift in the Royal Society Range over the past 7.8 Ma to less than 67 m, assuming present sea level as a datum. Similarities between the denudational and surface uplift histories of the Royal Society Range and the adjacent Dry Valleys area show that the latter has not experienced an unusual tectonic and glacial history, as has been previously suggested. Our analysis strongly supports the notion of a stable East Antarctic Ice Sheet and minimal landscape modification in the Royal Society Range since at least the mid-Miocene.

Quantifying denudation rates on inselbergs in the central Namib Desert using in situ–produced cosmogenic 10Be and 26Al

In situ-produced cosmogenic isotope concentrations in bedrock surfaces provide valuable estimates of site-specific, long-term rates of denudation and provide constraints for numerical landscape-evolution models. Measurements of cosmogenic {sup 10}Be and {sup 26}Al from graphite inselbergs in the arid to hyperarid central Namib Desert, Namibia, indicate a mean rate of summit lowering of 5.07 {+-} 1.1 m/m.y. over the past {ge} 10{sup 5} yr. The persistence of an arid climate in the region suggests that a similar rate may have prevailed for the past {approximately} 10 m.y. and possibly throughout much of the Cenozoic. Some samples have complex exposure histories that can be explained by the mode of inselberg weathering and mass wasting. The denudation rates estimated here are an order of magnitude higher than those reported for inselbergs in a significantly more humid environment in South Australia. This difference may largely be due to active salt weathering in the central Namib as a result of high levels of coastal fog precipitation.

Sub-aerial denudation of passive margins: regional elevation versus local relief models

Abstract Variations in denudation rates, which are an important factor in the modification of the sub-aerial topography of passive margins, have frequently been modelled as being proportionate to regional elevation. To be valid, this assumption requires that mean slope gradient and the related property of local relief show a strong positive relationship to regional elevation since slope angle is a primary determinant of denudation rate. Analysis of the relationships between these variables based on the 1:250,000 scale topographic map coverage of southern Africa demonstrates that regional (modal) elevation explains only 0.5% of the variance in mean sine slope gradient and only 0.14% of the variance in local relief (measured over 15′ × 15′ areas). These results demonstrate that regional elevation cannot be used to predict denudation rates for passive margins, such as those of southern Africa, where a dissected coastal zone is separated from a high interior plateau by a major escarpment. On the basis of a previous regression analysis relating local relief to sediment and solute transport by rivers the data for southern Africa indicate denudation rates of > 80 m Ma −1 over much of the coastal zone and −1 across most of the elevated interior. Extrapolation of these results over geological time, however, necessitates either assumptions about the local relief of palaeotopography or independent evidence of long-term denudation rates.