Fracturing on Lava Flow Surfaces
I. Crease Structures
Small crease structure on the Medicine Lake Dacite Flow showing striae on the fracture surfaces.
When a crusted lava flow spreads laterally, the flow surface may develop fractures called crease structures. These fracture surfaces spread outward in a manner analogous to the movement of divergent mid-ocean ridges or continental rift zone tectonic plates (Anderson and Fink, 1992). This lateral spreading of lava results in the concentration of tensile stress along a line oriented perpendicular to the direction of spreading. The cooled crust of the extrusion is therefore torn apart about this line of stress concentration, forming a central valley that may expose hot, ductile material from the flow interior to the atmosphere. As this ductile material rapidly cools upon surface exposure, it becomes part of the passive, rigid plugs (crease structure walls) that move apart from one another in response to the spreading of the underlying ductile flow interior. Fractures formed by this rifting process can be remarkably smooth and may occupy areas of many hundreds of square meters.
Large crease structure bisecting the June 1981 Mount St. Helens dome lobe. The lobe is approximately 70 meters wide. Photo courtesy of the USGS.
Entire domes or dome lobes may consist of single, large crease structures (Anderson and Fink, 1990; 1992). Individual crease structures may also occur locally in parts of a lava flow where buoyant vesicular magma rise from an internal flow zone to the surface (Fink, 1983). Crease structures are also commonly located on the crests of compressional ridges, and their orientation suggests they form from tensional stresses acting perpendicular to maximum compression (Anderson and Fink, 1992). These otherwise-smooth fractures are typically covered by long, step-like striae that cut across flow banding and reflect the episodic inward growth of the propagating crack. These features form in a manner analogous to striae found on columnar joint surfaces where each striation represents a single fracture increment/cycle.
II. Dilated Fractures on Inflated Basalt Flows
We have also studied dilated fractures in Hawaiian pahoehoe lava flows. We find that these fractures contain three zones that show the kinematics of inflation. An upper columnar zone forms through thermal contraction prior to inflation, the middle planar zone reflects inflation-induced tension, and the lower banded zone contains evidence of brittle and ductile deformation.
Left: Fracture in a tumulus displaying an upper columnar zone (extending ~30cm down from the flow surface), underlain by a middle planar zone of similar thickness, and a banded basal zone extending to the base of the fracture. Brandi Wood, seen here doing the important work, was a senior at Black Hills State University when she became involved in this project. She spent nearly 2 weeks in Hawaii, working long, extremely hot days as part of a grant she received from the BHSU Nelson Scholarship Committee.
The formation of the lower banded zone requires varying strain rates during fracture propagation, and is best explained by a model where small pulses of lava inject beneath the cooled flow crust through a network of preferred pathways. We believe that this inflation mechanism is incapable of producing areally extensive continental flood basalts on Earth, although it can explain related features on large Martian volcanoes. We recently published this work and a reply to a comment on this work by Self and others in Earth and Planetary Science Letters (Anderson et al., 1999; 2000).
III. Fracturing in Continental Flood Basalt Interiors
I have been investigating fracture types and patterns in the interiors of continental flood basalts in the Columbia River Plateau with Ellen Stofan, Sue Smrekar (both at JPL) and John Guest (University College, London). Fractures originate in response to stress/strain regimes within the flows, thus can provide important clues regarding the emplacement of these enormous flows. Currently, there is debate regarding the rate of flood basalt emplacement. Historically, it's been held that these extensive flows formed quickly as "floods" of lava inundated a large area. Recently, others have suggested that flood basalts formed much more slowly, similar to how Hawaiian pahoehoe sheet flows are emplaced. We are analyzing the fracturing within these flows to determine whether the stress/strain relationships favor either of these models.