Bounding surfaces in aeolian dunes

Bounding surfaces are an integral part of aeolian dune stratigraphy.

Aeolian dune stratification is characterized by crossbedding. Crossbeds occur in sets and cosets that are three-dimensional structures bound by erosional surfaces. A coset consists of two or more similar crossbed sets stacked vertically or amalgamated laterally (from the terminology of McKee and Weir, 1953). The surfaces of a crossbed set define the limits of cross-stratification (foresets); they can be conformable or discordant with respect to the cross-stratification, but inevitably are disconformable (erosional) with respect to other crossbed sets.

Bedforms at all scales in aeolian dune systems are rarely static. They migrate, respond to changes in wind direction and strength, override other bedforms, or erode earlier-formed deposits. Sand dunes stabilized by vegetation or evaporite crusts can be reactivated. The resulting stratigraphic architecture will contain a multitude of crossbeds bound by surfaces that reflect these processes (e.g., Bristow and Mountney, 2013).

Brookfield (1977) was one of the first to attempt a hierarchical classification of aeolian bounding surfaces (Brookfield and Silvestro, 2011). He defined 1^st, 2^nd, and 3^rd order surfaces based on their lateral extent and geometry, and cross-cutting or discordant relationships, all of which reflect the migratory behaviour of dunes. Brookfield defined:

Third-order surfaces (the smallest) enclose single crossbed sets and arise from local changes in wind direction and velocity.
Second-order surfaces represent the migration of smaller dunes over larger dunes.
First-order surfaces are the most extensive and represent the downwind passage of large dunes (e.g., draas , barchan dunes). First-order surfaces cut across 3^rd and 2^nd order surfaces.

Brookfield’s classification was modified by Kocurek (1981, 1988, 1996 Chapter 5), who replaced the numbered hierarchy with descriptive terms that reflect the basic processes involved in their formation: reactivation, superposition, and interdune surfaces. The changes in terminology were based in part on numerical modeling experiments by Rubin and Carter (1987) and Rubin and Hunter (1983). Their models predicted significant degrees of complexity in the kinds of erosional surfaces resulting from changes to lee face and crest line geometry, variable wind directions (including wind backflow and deflection by the dunes themselves), and erosion. It is also possible for the surfaces to be superimposed, for example where a 3^rd order surface erodes part of a 2^nd order surface. The three bounding surfaces correspond in general aspects with Brookfield’s scheme; the two classification schemes have been used interchangeably for several years.

Note that bounding surfaces are observed structures in sedimentary rocks but their designation as reactivation surfaces and so on is fundamentally interpretive.

Reactivation surfaces

These surfaces represent the reactivation of a dune lee-face following a change in wind direction or velocity. In this case the bedform migration direction and rate change but the bedform itself remains largely intact. Surface manifestation can be as subtle as a minor increase or decrease in foreset dip, or a more dramatic discordance that truncates crossbed foresets. These surfaces are ubiquitous because wind dynamics are continually changing, even with prevailing winds. Reactivation surfaces are also common on subaqueous, tidally influenced bedforms.

Superposition surfaces

Superposition surfaces develop where smaller bedforms migrate over earlier-formed dunes or large ripples. The contact between the two bedform generations is usually discordant where erosion occurs in down-wind scour troughs. The most common example observed on nearly all sand dunes (on Earth and Mars) is the small wind ripples that populate the stoss and lee faces of larger dunes. Superposed bedforms can migrate in the same direction as the underlying structure (down the lee face) or at high angles, for example where dunes or ripples migrate along or up the lee face.

Kocurek (1996, Op. Cit.) maintains that superposition generally takes place on the lee faces of larger dunes. This begs the question – how do we distinguish between reactivation and superposition surfaces if both reflect lee face processes? Bristow (2019) attempts to answer this question using a modern example of barchan dune morphology and GPR imagery in the Tarfaya dune system, Morocco. The ground penetrating radar (GPR) profile (parallel to the slip face) shows several bounding surfaces where lee face foresets are truncated. Are these reactivation or superposition surfaces?

Bristow observed superposed, low-amplitude bedforms that migrated up the barchan dune stoss face; bedforms approaching the dune crest developed a slip face and scour trough immediately downwind. The scoured surface in the GPR image had truncated the barchan dune brink and lee-face foresets. The superposed bedforms have wavelengths averaging about 20 m and migration rates up to 2m/day, nearly 30 times the estimated migration rate of the main barchan dune. This implies that superposition surfaces arising from smaller bedforms should be common in the dune stratigraphic record. We can reasonably infer that the older bounding surfaces in this profile also formed as superposition surfaces but in fact we cannot logically exclude reactivation processes based on the 2D profile alone.

Kocurek (1996, op. Cit.) maintains that we can distinguish between reactivation and superposition surfaces in sections parallel and perpendicular to dune migration, where surface dip directions can be measured. Reactivation surfaces tend to be subparallel to foreset dips; superposition surfaces at greater angles. But the true dips of these surfaces can only be established from 3-dimensional exposures.

Interdune surfaces

Interdune areas that separate large dunes range in lateral extent from relatively narrow scour troughs downwind of advancing dune lee faces, to broad sand sheets (Kocurek, 1981, Op. Cit.). They tend to be areas of low sediment accumulation. In dry dune systems they are commonly floored by coarse-grained deflation lags, desiccation structures, and rippled sand. The depth of erosion across dry interdune flats generally corresponds to the depth of scour in front of large migrating dunes. In wet dune systems, the depth of erosion is limited by the depth to the watertable. Wet interdune flats may be vegetated. Evaporite crusts can also form in areas that have access to saline groundwater or incursions by seawater (e.g., sabkhas). Peats can accumulate in interdune areas that become ponded.

Interdune bounding surfaces represent the migration and preservation of large dunes over interdune deposits. In systems that accumulate sand, the surfaces must climb downwind. However, the angle of climb is low so that the surfaces appear flat in outcrop. They are distinguished from reactivation and superposition surfaces by their lateral extent, deflation lags, desiccation structures, bioturbation, and soils, peats, and root structures in wet systems.

Overlapping bounding surfaces

Measurements of dune migration rates indicate that smaller bedforms generally move faster than larger dunes (they require less sand to regenerate their bedform). Large dunes tend to move relatively slowly – a few metres per year for some of the largest Draas and compound dunes, and 20-50 m/year for smaller, simpler structures like barchan and parabolic dunes; migration rates exceeding 70m/year have also been recorded.

In contrast, smaller superposed bedforms commonly migrate several metres/day – one and two orders of magnitude faster than the parent dune they override. Reactivation surfaces that depend on changing wind dynamics also have the potential to imprint a record several times a day.

Thus, a year’s worth of dune migration over interdune deposits will also produce many superposition and reactivation surfaces that are potentially preserved in the dune stratigraphy. The superposition surfaces may truncate or overlap earlier-formed reactivation surfaces, and superposition surfaces may be reactivated by frequently changing wind dynamics. It is also possible for superposition surfaces to overlap an interdune surface as the larger dune migrates.

We can observe bounding surface processes in modern aeolian dune systems; we can also predict possible stratigraphic architectures by modeling the interactions of dunes to changing environmental conditions. However, transferring this knowledge to the rock record, and to outcrop or regional scale problems is nontrivial. Confidence in our assignations of bounding surfaces must also be influenced by the fact that most exposures of ancient dune successions are 2-dimensional. These problems are nicely illustrated in some outcrop examples from the Pleistocene of west Auckland (NZ).

The traced crossbed sets and their bounding surfaces in this 2D outcrop have been numbered according to their relative age of formation (1 is oldest). The primary direction of dune migration is to the right based on additional outcrop information. If we interpret the bounding surfaces based on the definitions alone, without considering possible variations in bedform orientation, then surfaces 3, 4, 5, 6, 7, 9, and 10 are classified as superposition surfaces; surface 12 may be a scour trough associated with the migration of a small bedform. Number 11 and the upper part of 5 are good candidates for reactivation surfaces; additional, subtle reactivation surfaces also occur within these crossbed sets. The upper part of 8 also appears to be a reactivation surface except for the intervening crossbed sets bound by surfaces 9 and 10.

Crossbed sets bound by surfaces 4, 5, 6, 7, 9, and 10 may be candidates for the superposition category. But what if these are strike sections of crossbed sets that formed at high angles to the primary migration direction. In other words, if the relatively flat lying foresets are viewed from either up- or down-current, then the bounding surfaces could be reactivation surfaces. We cannot distinguish between the two possible bounding surfaces unless we can view these deposits in the 3^rd dimension.

Super surfaces

Super surfaces (e.g., Kocurek, 1988, Op.Cit.) represent the termination of entire or large parts of aeolian dune systems (ergs, coastal dune systems). The cessation of dune activity over an entire system (ergs, coastal dunes) requires significant changes in climate, sand supply, sea level, watertable elevation, and tectonics. None of these broad environmental conditions are mutually exclusive. For example, a change in climate from wet to dry will reduce the chemical weathering of bedrock and sediment routing and supply. A rise in sea level may truncate coastal dunes or move the dune system inland (or both, depending on the relative rates of shoreline migration and the ability for the dune system to reestablish itself. Regional watertables will also rise in concert with sea level, resulting in a greater propensity for ponding and stabilizing vegetation.

Super surfaces are stratigraphic discordances of regional extent. Surfaces exposed for long periods will develop paleosols – organic soils, evaporite crusts, caliches, or peats where there is significant flooding or ponding. Coastal dune systems truncated by sea level rise may show ravinement surfaces. Super surfaces may also develop erosional relief during subsequent sea level fall and reestablishment of coastal drainage. An example from a Pleistocene coastal dune system is shown at the top of this page.