Citation for published item:idmondsD wF nd rumphreysD wFgFF nd ruriD iF nd rerdD F nd dgeD qF nd wsonD rF nd veddenD F nd lilD wF nd frlyD tF nd eiuppD eF nd ghristopherD F nd qiudieD qF nd quidD F @PHIRA 9reEeruptive vpour nd its role in ontrolling eruption style nd longevity t oufri ere rills olnoF9D wemoirsFD QW F ppF PWIEQISF Further information on publisher's website: Use policyThe full-text may be used and/or reproduced, and given to third parties in any format or medium, without prior permission or charge, for personal research or study, educational, or not-for-prot purposes provided that:• a full bibliographic reference is made to the original source • a link is made to the metadata record in DRO • the full-text is not changed in any way The full-text must not be sold in any format or medium without the formal permission of the copyright holders.Please consult the full DRO policy for further details. Schmidt and Poli, 1999). The generation of mafic magmas in the "hot 52 zone" at the base of the arc crust is fundamentally controlled by the H 2 O content of the melts, 53 which act to "flux" the amphibolite crust and cause assimilation (Annen et al,. 2006). In 54 shallow storage areas, the exsolved fluid content of stored magma directly influences the 55 compressibility and hence response of the magma body to changes in pressure or volume, 56 ultimately determining magma ascent rates and eruption style, as well as eruption duration 57 and size (Huppert and Woods, 2002). The rate of crystallisation of magma in crustal 58 reservoirs is influenced by the opposing effects of fluxing by CO 2 -rich gases from depth, 59 which act to "freeze" the magma (Blundy et al., 2010), and new batches of incoming hot, 60 volatile-rich magma, which act to "defrost" it, by heating and transferring H 2 O-rich fluids 61 (e.g. Bachmann and Bergantz, 2006). Sulphur partitions into this fluid (e.g. Scaillet et al., 62 1998;Zajacz et al., 2012), further enhancing the proportion of vapour coexisting with the 63 magma prior to the eruption, and acting as a convenient tracer to measure in volcanic gases 64 (e.g. Gerlach et al., 2008). In the conduit, transitions between lava dome building and 65 explosive Vulcanian activity are controlled by finely-balanced feedbacks involving the 66 development of permeability during degassing, and the increase in viscosity and consequent 67 retardation of bubble growth caused by crystallisation and H 2 O exsolution from the melt 68 (Melnik and Sparks, 1999;Sparks et al., 2000;Melnik and Sparks, 2002; Clarke et al., 2007). 69 CarmichaelThe arc magma system is complex; petrological and geochemical evidence points to the 70 erupted porphyritic andesite being a hybrid, the result of perhaps countless recharge events by 71 internal energy associated with dissolved volatiles can be released into the magma chamber, 87 and this is a mechanism for inducing complex eruption cycles on long timescales for volatile-88 rich magmas (Huppert and Woods, 2002). One mechanism for this is the presence of a large 8...

An exceptional opportunity to sample several large blocks sourced from the same region of the growing Soufrière Hills lava dome has documented a significant increase in the presence of mafic enclaves in the host andesite during the course of a long‐lived eruptive episode with several phases. In 1997 (Phase I) mafic inclusions comprised ∼1 volume percent of erupted material; in 2007 (Phase III) deposits their volumetric abundance increased to 5–7 percent. A broader range of geochemically distinctive types occurs amongst the 2007 enclaves. Crystal‐poor enclaves generally have the least evolved (basaltic) compositions; porphyritic enclaves represent compositions intermediate between basaltic and andesitic compositions. The absence of porphyritic enclaves prior to Phase III magmatism at Soufrière Hills Volcano suggests that a mixing event occurred during the course of the current eruptive episode, providing direct evidence consistent with geophysical observations that the system is continuously re‐invigorated from depth.

Citation for published item:lilD wF nd frlyD tF nd rumphreysD wFgFF nd idmondsD wF nd rerdD FeF nd ghristopherD FiF @PHIRA 9ghrteriztion of m( enlves in the erupted produts of oufri ere rills olnoD wontserrtD PHHW to PHIHF9D wemoirsFD QW F ppF QRQEQTHF Further information on publisher's website: he iruption of oufri ere rills olnoD wontserrt from PHHH to PHIHF ghpter IVF Use policyThe full-text may be used and/or reproduced, and given to third parties in any format or medium, without prior permission or charge, for personal research or study, educational, or not-for-prot purposes provided that:• a full bibliographic reference is made to the original source • a link is made to the metadata record in DRO • the full-text is not changed in any way The full-text must not be sold in any format or medium without the formal permission of the copyright holders.Please consult the full DRO policy for further details. et al. 2003). 101The presence of mafic enclaves in SHV andesite is ascribed to the interaction between 102 mafic magma and the andesitic host magma, which is perhaps the trigger and driver for the Methods 129Samples of andesite and mafic enclaves were collected from a wide range of locations around 130SHV from deposits emplaced during Phase V activity ( proportion of sieved inherited phenocrysts in different enclaves (see Fig. 3). 195 Results 196The andesite erupted in Phase V of the eruption is porphyritic with a fine-grained 197 groundmass, and contains mafic enclaves as in earlier phases ( Fig. 1) (Murphy et al. 1998(Murphy et al. , 198 2000Harford et al. 2002; Barclay et al. 2010). Some andesite blocks contain distinctive 199 streaked highly crystalline layers of amphibole and plagioclase. Pumice is porphyritic with a 200 fine-grained groundmass and often contains mafic enclaves. 201Total measured mafic enclave abundances within andesitic Phase V blocks range 202 from 2.9% to 8.2% from point counting, with a mean of 5.6% (Table 2). The size of 203 individual enclaves ranges from 1 to 80 cm; however, ∼95% of the enclaves were <10 cm in 204 apparent diameter (Fig. 2). We categorised Phase V enclaves into three broad types that were 205 readily identifiable in the field using characteristics such as phenocryst proportions, the 206 nature of the margin between enclave and andesite, vesicularity, enclave size and shape, and 207 groundmass size and colour (Table 3). The classification scheme applied by Barclay et al. 208(2010) is insufficient to describe the large textural diversity of the Phase V enclaves. 209Type A enclaves are characterised as phenocryst-poor, vesicle-rich, with dark grey 210 groundmass and chilled margin (Table 3, Fig. 1). In the field these enclaves are readily 211 identified by their dark grey colour caused by the fine-grained groundmass composition. (Table 217 2). These have the smallest mean diameter of all the enclave types measured (2.3 cm), 218although large enclaves over 18 cm were also measured (Fig. 2). 219Type B enclaves are characterised as phenocryst-rich, vesicle-po...

Mafic enclaves hosted by andesite erupted at the Soufrière Hills Volcano between 1995 and 2010 yield insights into syn-eruptive mafic underplating of an andesite magma reservoir, magma mixing and its role in sustaining eruptions that may be widely applicable in volcanic arc settings. The mafic enclaves range in composition from basalt to andesite and are generated from a hybrid thermal boundary layer at the interface between the two magmas, where the basalt quenches against the cooler andesite, and the two magmas mix. We show, using an analytical model, that the enclaves are generated when the hybrid layer, just a few tens of centimetres thick, becomes buoyant and forms plumes which rise up into the andesite. Mafic enclave geochemistry suggests that vapour-saturated basalt was underplated quasi-continuously throughout the first three eruptive phases of the eruption (the end member basalt became more Mg and V-rich over time). The andesite erupted during the final phases of the eruption contained more abundant and larger enclaves, and the enclaves were more extensively hybridised with the andesite, suggesting that at some time during the final few years of the eruption, the intrusion of mafic magma at depth ceased, allowing the hybrid layer to reach a greater thickness, generating larger mafic enclaves. The temporal trends in mafic enclave composition and abundance suggests that basalt recharge and underplating sustained the eruption by the transfer of heat and volatiles across the interface and when the recharge ceased, the eruption waned. Our study has important implications for the petrological monitoring of long-lived arc eruptions