9Conceptual models for the evolution of dilatant faults in volcanic rift settings involve a step-wise 10 growth pattern, involving upward propagation of subsurface faults, surface monocline formation, 11 which are breached by subvertical, open faults. Immature, discontinuous normal faults are 12 considered representative of the early stages of mature, linked faults that accommodate extensional 13 strains. We consider the evolution of surface-breaking normal faults using a comparison of the 14 distribution and geometry of normal faults from two volcanic rift zones: the Koaʻe fault system, 15Hawaiʻi, and the Krafla fissure swarm, NE Iceland. Field mapping highlights similarities to current 16 predicted geometries, but also prominent differences that are not reconciled by current models. 17Variable deformation styles record magma supply changes within the rift zones, which drive local 18 strain rate gradients. Building on existing studies, we present a conceptual model of fault growth 19 that accounts for spatial and temporal changes in strain rate within the deforming regions. We 20propose that faults in separate rift systems may not advance through the same stages of evolution 21and that faults within individual rift systems can show differing growth patterns. Variations in 22 2 surface strains may be indicative of subsurface magmatic system changes, with important 23 implications for our understanding of volcano-tectonic coupling. 24
25Key words: normal fault; monocline; extension; basalt; volcanic rift 26 27
Introduction 28Normal fault systems comprise discontinuous, non-collinear segments, with overlaps and segment 29 linkage commonly resulting in characteristic overlapping or step-like geometries across a broad 30 range of scales (e.g. Segall and Pollard, 1980;Peacock, 2002; Long and Imber, 2011). Regional 31 extension is conserved ahead of first-order fault terminations by areas of folding and linking faults 32 and fractures (e.g. Morley et al., 1990;Faulds and Varga, 1998). The geometry and distribution of 33 structures within these domains play an important role in the tectono-stratigraphic development of 34 rift basins (e.g. Lambiase and Bosworth, 1995; Sharp et al., 2000; Hus et al., 2006), and the 35 evolving fluid flow properties of fault zones (e.g. Manzocchi et al., 2010;Seebeck et al., 2014). 36Much of our current understanding of the growth of normal fault populations and fault zone 37 architecture is derived from studies of faults in clastic sequences using combinations of: (1) fault 38 analysis and scaling relationships, based on field and seismic data-derived measurements of 39 displacement and length versus width (e.g. Ferrill and Morris, 2001;Peacock, 2002; Walsh et al., 40 2003;Nixon et al., 2014); (2) scaled-analogue modelling (e.g. Holland et al., 2006; Tentler and 41 Acocella, 2010); and (3) numerical-based modelling techniques (e.g. Crider and Pollard, 1998; 42 Maerten et al., 2002;Schöpfer et al., 2006). Many of these studies have focussed on fault 43 propagation and se...