New insights into the mechanics of sill emplacement provided by field observations of the Njardvik Sill, Northeast Iceland

Abstract

Sills are concordant sheet-like bodies of magma. Their mechanics of emplacement is an important but still not fully understood topic. The well-exposed basaltic Njardvik Sill in the extinct Tertiary Dyrfjöll Volcano in Northeast Iceland offers exceptionally clear insights into the mechanism of sill emplacement. The sill is multiple and consists of at least 7 units (sills) all of which were emplaced along a sharp contact between a rhyolitic intrusion and adjacent basaltic lava flows. Each sill unit was supplied with magma from an inclined sheet. The contacts between the sheets and the sill units are very clear and show that the sill units are much thicker than their feeder sheets. Since the Njardvik Sill consists of separate units, it obviously did not evolve into a homogeneous magma body. Nevertheless, the abrupt change in dip and thickness from inclined sheets to horizontal sills at this particular locality indicates that the earlier sills were influencing the stress field in their vicinity during the subsequent sheet injections. The local stresses around the newly formed sill units forced each of the subsequently injected sheets to change into sills. The Njardvik Sill can be followed laterally in a coastal section for 140 m until it ends abruptly at a fault that cuts the sill. Using these field observations as a basis, a numerical model shows how an inclined sheet opens up the contact between the felsic intrusion and the basaltic lava pile, along which the sill emplacement takes place. The results suggest that sill emplacement is primarily the result of stress rotation at contacts between layers of contrasting mechanical properties. There, the orientation of the maximum principal compressive stress σ1 is horizontal. Hence, such contacts can represent interfaces along which sill emplacement is encouraged.Once a sill has been emplaced, it extends the stress field with a horizontal orientation of σ1. Consequently, inclined sheets and dykes injected near the sill will be deflected into sills. The injection frequency of further sill units controls if the sill can grow into a larger magma body by mixing of the newly supplied with the initially injected magma. In case of the Njardvik Sill, the injection frequency was low, so subsequently emplaced sill units can be distinguished. The contacts between the sheets and the sill units are very clear and show that the sill units are much thicker than their feeder sheets. Since the Njardvik Sill consists of separate units, it obviously did not evolve into a homogeneous magma body. Nevertheless, the abrupt change in dip and thickness from inclined sheets to horizontal sills at this particular locality indicates that the earlier sills were influencing the stress field in their vicinity during the subsequent sheet injections. The local stresses around the newly formed sill units forced each of the subsequently injected sheets to change into sills. The Njardvik Sill can be followed laterally in a coastal section for 140 m until it ends abruptly at a fault that cuts the ...

What Controls Sill Formation: An Overview From Analogue Models

“…Table shows the range of values for each parameter in the models and nature, with the associated scaling factor. The scaled Young's modulus considered in our models corresponds to 10 10 Pa, within the range of natural rocks (10 9 –10 10 , Kavanagh et al, ; Table ), while the average intrusion lengths ( H ) before dike to sill rotation in the experiments (~11 cm) corresponds to ~2,200 m in nature, consistently with the length of dikes feeding sills (Burchardt, ; Famin & Michon, ). The lateral compression imposed in the experiments corresponds, scaled, to 10 6 Pa (Table ), thus within the range of compressive stresses in nature (between 10 5 and 10 8 Pa; Menand et al, ).…”

“…Natural examples of rigidity contrasts (i.e., stiff layer above a weak one, as imposed in our models) are common in many volcanic areas with alternating soft scoria (or pyroclastic) layers and stiff lava flows, such as in SW Iceland and Tenerife (Gudmundsson, ; Gudmundsson & Løtveit, ). Field observations on the Njardvik Sill (Northeast Iceland; Burchardt, ) show initial sill emplacement when an inclined sheet dipping 46° met the horizontal contact between a rhyolitic intrusion of the Njardvik Silicic Complex and the overlying basaltic lava flows. For the Njardvik Sill, the rigidity contrast between the rhyolitic Njardvik Silicic Complex and the basaltic lava flows (inferred to be E u / E l = 3) might have caused a local stress rotation and the mechanical properties of the interface between rhyolite and basalt might have encouraged stress rotation and the intrusion of the first sill unit of the Njardvik Sill (Burchardt, ).…”

Rapid laccolith intrusion driven by explosive volcanic eruption

“…4a). Lateral injection of magma to build the initial sill may occur along subhorizontal weak planes2425 such as those between lavas and pyroclastic deposits beneath the vent (Supplementary Fig. 3).…”

Application of open‐source photogrammetric software MicMac for monitoring surface deformation in laboratory models

“…Although commonly used, these models have serious limitations: They model simplistic intrusion and fault shapes, such as point source [ Mogi , ; Masterlark , ], tensile [e.g., Okada , ; Amelung et al ., ; Wright et al ., ; Chang et al ., ; Sigmundsson et al ., ], or shear dislocation rectangle [ Okada , ; Shen et al ., ; Gusman et al ., ; Yue et al ., ]. These deformation sources are not representative of the complex shapes of magmatic intrusions or fault planes in nature [e.g., Burchardt , ; Lohr et al ., ; Burchardt et al ., ]. They model static intrusions/fault planes, such that they do not account for the complex magma propagation mechanisms [ Mathieu et al ., ; Abdelmalak et al ., ] or fault mechanics [ Mair and Abe , , ; Brodsky and Lay , ]; It is impossible to quantify the uncertainties of the model results and so to test their robustness. The main reason is that active geological processes occur in the subsurface, so that the results of the modeling cannot be validated by direct observations. …”