What Drives the Lateral Versus Vertical Propagation of Dikes? Insights From Analogue Models

Abstract

Volcanic eruptions are usually fed by dikes. Understanding how crustal inhomogeneities and topographic loads control the direction (lateral/vertical) and extent (propagation/arrest) of dikes is crucial to forecast the opening of a vent. Many factors, including buoyancy, crustal layering, and topography, may control the vertical or lateral propagation of a dike. To define a hierarchy between these factors, we have conducted analogue models, injecting water (magma analogue) within gelatin (crust analogue). We investigate the effect of crustal layering (both rigidity and density layering), topography, magma inflow rate, and the density ratio between host rock and magma. Based on the experimental observations and scaling considerations, we suggest that rigidity layering (a stiffer layer overlying a weaker one) and topographic gradient favor predominantly lateral dike propagation; inflow rate, density layering, and density ratio play a subordinate role. Conversely, a softer layer overlying a stiffer one favors vertical propagation. Our results highlight the higher efficiency of a stiff layer in driving lateral dike propagation and/or inhibiting vertical propagation with respect to the Level of Neutral Buoyancy proposed by previous studies.

What Controls Sill Formation: An Overview From Analogue Models

“…Merging our and Urbani et al's () results, we summarize which parameters influence sill emplacement, lateral dike propagation and vertical dike propagation (Figure ).…”

“…Therefore, we consider both the high rigidity contrast and the high buoyancy pressure as first‐order parameters affecting sill formation. Indeed, in Experiments 24, 25 and 26, where E u / E l = 1.8, we observe dike arrest and lateral dike propagation (with the dikes showing a higher width and lower length compared to other experiments; Figure c; Urbani et al, ) because the injected water was denser compared to the other experiments with the same E u / E l ( ρ f = 1,116.4 kg/m 3 ); therefore, the buoyancy pressure was not sufficient to overcome the rigidity contrast and promote sill rotation ( P m = 8.2 Pa corresponding to ~0.05–1.1 MPa in nature).…”

“…Both E u / E l = 1.8 and a high buoyancy pressure ( P m ≥ 60 Pa in our case) are necessary and sufficient conditions for sill emplacement. If a dike has no sufficient buoyancy pressure to intrude the interface or become a horizontal sill, it arrests its vertical propagation and propagates predominantly laterally as a blade‐like dike (Kavanagh et al, , ; Urbani et al, ). Our experiments suggest that the critical buoyancy pressure to have sills is P m ≥ 60 Pa; even though it is of the same order of magnitude, this is slightly higher than that previously proposed (i.e., ~10 Pa; Kavanagh et al, ).…”

“…Main parameters controlling sill emplacement, lateral and vertical dike propagation. * Urbani et al (, ).Figures show the section view.…”

“…Topographic unloading (Corbi et al, ; Maccaferri et al, ) or loading (Pinel & Jaupart, ; Watanabe et al, ) may locally alter the crustal stresses, favoring sill emplacement. The ratio between the thickness of an upper and lower layer may affect lateral dike or sill propagation (Urbani, ). Finally, magma cooling (Taisne & Tait, ) may form sills only if solidification effects are moderate, since high cooling rates induce dike arrest, while low cooling rates promote dike propagation (Chanceaux & Menand, , ).…”