Heat transfer in active volcanoes was investigated in this thesis. A general model of a crater lake system was developed that takes inputs of lake temperature, volume, chemical content, and meteorological conditions, and outputs the mass, energy, and chemical flows to and from the lake. The model was applied to lakes at Pods (Costa Rica) and Ruapehu (New Zealand), yielding volcanic power outputs of -102-103 MW, and heat fluxes of -102-104 W m-2. Heat is added to the lakes by hot brine and steam, derived from lake seepage and magmatic gas. The heat source is magma crystallising, cooling, and degassing. Background heat inputs are maintained by hydrothermal infiltration of magma, releasing latent and specific heat. Infiltration of the conductive boundary layer surrounding magma was modelled. The permeability created by contractive cooling was equated with the permeability required for two-phase convection to transport heat away from the boundary. Infiltration rates of -1-100 m a-1 (metres/year), into conductive layers -30-0.3 m thick, creating permeabilities of _10-10-10-14m 2, will provide the required heat flux. Cracking temperatures of magma depend upon infiltration rate, ranging from hydrothermal system temperatures at slow rates, to magma temperatures at the fastest rates. Predicted maximum rates are -300 m a-1 for near-surface magma, and -800 m a-1 for magma at -1-3 km depth. Measured S02 fluxes at Pods, and calculated influxes of HCl to both lakes imply that degassed magma volumes (-0.004-0.08 km3 a-i) are much larger than likely intrusions. A new model was developed of small, vesiculating intrusions that circulate magma due to the density increase associated with gas loss. Dense, degassed magma descends, whilst buoyant, volatile-rich magma rises from a deep source. Pipe-like intrusions of radius -5 m, tapping magma volumes >-0.05 km3, can produce the gas fluxes needed. Intrusions of this type probably occurred in 1980/81 and 1986 at Pods, and in 1968,1971,1975,1977,1981 and 1985 at Ruapehu, and were followed by intermittent eruptions and degassing. This degassing mechanism probably occurs at many volcanoes where high gas fluxes are observed, but no evidence exists for large, shallow intrusions. A model of compressible fluid flow in a rough fumarole conduit, with conductive heat loss to the surroundings, allows fumarole temperatures to be used to estimate the depth of their magma source. This also indicates shallow magma was emplaced at Pods in 1980/81 and 1986. In summary, heat transfer is achieved by a combination of intermittent gas release from minor shallow intrusions, together with infiltration of deeper magma. Infiltration is one mechanism for providing fractures allowing the release of gas from shallow intrusions, and circulation probably ceases due to freezing caused by infiltration.

The Dam Feliciano Belt is a Pan-African mobile belt from the Ribeira orogen of southern Brazil. A detailed field and geochemical traverse along the BR392 road section between Pelotas and Cacapava do SuI identifies two major tectonic domains; the Pelotas Batholith and the Santana Metamorphic Belt, striking NNE-SSW parallel to the major foliation of the belts. The two belts are separated by a Triassic basin with flat lying red beds and interbedded andesites and rhyolites. The Santana Metamorphic Belt is a NW-verging fold belt with a metamorphosed shelf sequence of quartzites, marbles and graphitic schists and a polydeformed Lower Proterozic gneissic basement deformed during the Pan-African orogeny. Detailed mapping recognized four phases of deformation in the basement gneisses, three of which are recorded in the basement schists and cover sequence. There is also evidence of late NE-verging thrusting post-dating the formation unmetamorphosed Paleozoic sediments. Late extension caused NW-SE and NE-SW normal faulting. Metamorphism occurred contemporaneously with D2 and 03, and PT conditions for peak: metamorphism have been calculated as 8.6Kb and 60QOCwithin the basement schists corresponding to garnet growth during D3 deformation. Three phases of granite intrusion are recognized in the Santana Metamorphic Belt The Santana Granite (8oom.y.) represents the first phase intruding the basement and it is folded by 02 and 03. The Campinas Granite (Soom.y.) has a 03 foliation and the Cacapava Granite (474m.y.) is post-tectonic and intrudes the NW portion of the Santana Metamorphic Belt The Pelotas Batholith is almost entirely composed of granitoids of Pan-African age (600- 4S0m.y.). Both D2 and D3 are recognized in the batholith. A two fold subdivision based on geoc~emical and field criteria distinguishes the following categories of granitoid; foliated granitoids and unfoliated granites.

H2O and CO2 concentrations in olivine-hosted melt inclusions, assuming vapor saturation at the time of trapping, can be used to estimate crystallization depths for the olivine host. Estimating the true CO2 in melt inclusions is difficult, as much is lost to shrinkage bubbles, which form upon post-entrapment cooling and crystallization. Reheating olivine to temperatures above the melt inclusion trapping temperature and then quenching rapidly can restore CO2 to the glass because the CO2 in the bubble redissolves at high temperature. Previous work has established that olivine crystallization for the1959 Kilauea Iki eruption took place in the shallow summit reservoir, but crystallization depths have not been established for the rift extension of the eruption, at Kapoho. The new data presented here suggest that the most primitive Kilauea Iki component bypassed the summit reservoir for the east rift zone prior to the start of the eruption and was later erupted at Kapoho.

There has been active volcanism over a large part of western Kenya for the last 25 my. This has been accompanied by the development of the Kenyan Dome and Rift. The volcanics can be considered in terms of several space/time groupings based on their age and geographic position relative to the site of the present day rift valley. There is considerable variety in the petrography and chemistry of the volcanics. By examination of the more basic volcanics (> 4 wt. % MgO) it can be demonstrated that these variations are systematic through both time and space. As the alkalinity of the volcanics increases there is a sympathetic increase in the abundance of incompatible trace elements. Within the rift zone there is a decrease in alkalinity through time, and during the Miocene and Pliocene a north to south increase in alkalinity. Volcanics to the East and West of the rift zone are more alkaline than contemporaneous volcanics within the rift zone. The chemistry of the Kenyan Basic Volcanics is comparable to that of volcanics from ocean islands. North of the Kenyan Province, but associated with the same rift system, are the Ethiopian Volcanics, the chemistry of which is more akin to continental flood basalts. It is demonstrated that the chemical variety displayed by the Kenyan Volcanics can be produced by variable degrees of partial melting of and homogenous source mantle, and that this source must be garnet Iherzolite containing carbonate phases. The source must be enriched in Nb relative to other incompatible trace elements, but is not necessarily enriched in the light REE relative to the heavy REE. There may be variations in the Ba, P and Sr contents of the source. The trace element abundances of the source are similar to the source tapped by ocean islands. However, the ratios La/Nb and K/Nb show that unlike ocean islands the parental magmas to the Kenyan Volcanics are probably contaminated during transit through the sub-continental lithosphere. Geophysical and geotectonic evidence suggest the presence of a hotspot operating beneath Kenya. The variations in the chemistry, however, require a more sophisticated melting regime than that which would be produced a simple hotspot. It is possible that the volcanism is the result of partial melting in more than one melting regime and that these melting regimes have or have had a strong linear element to them, possibly related to upwelling of the mantle caused by passive stretching of the lithosphere.

Phreatomagmatic volcanoes form when ascending magma explosively interacts with surface or groundwater at shallow depths. Three types of phreatomagmatic activity are recognised- phreaticp phreatomagmatic (s. s. ) and surtseyan - based on the degree of involvement of magma with water and the depth of the interaction. Phreatic maars and phreatomagmatic tuff-rings are underlain by pipe-like diatremes but these structures are poorly developed or absent in surtseyan tuff-rings. Comparisons of phreatomagmatic volcanoes with their eroded diatreme equivalents, which contain subsided subaerially-deposited material, allow a model for activity of this type to be constructed. The Saefell tuff-ring SW Iceland is a surtseyantype structure whose crater remained open to the sea during most of its activity, allowing easy access of water to the magma. Base-surges sourced partly from directed blasts, formed large dunes with internal structures indicating deposition by density currents whose flowpower decreased with time and with distance from the vent. Syndepositional slumping and minor en masse collapse of crater deposits formed a pile of massive tuffs above which subsequent surge and airfall activity deposited a nestedg inner crater rim. The Medano tuff-ring Tenerife, is a phreatomagmatictype structure whose crater contains reworked tuffs deposited during subsidence into the underlying diatreme. Initial activity ejected much country rock material as magma contacted groundwater at depth but with time eruptions became more strombolian, as water was used up or failed to gain access to the vent. Surges were less common than in the Saefell eruption because the Medano water: magma ratio and explosion depth less often fulfilled the optimum conditions for surge production. The East Lothian diatremes in Scotland are subdivided into two groups on the basis of their infilling. The Red group diatremes contain high proportions of sediment and represent the subsided products of phreatic maars which erupted into a pile of water-rich poorly-consolidated alluvial plain sediments. The later Green group diatremes contain mainly juvenile basalt fragments and formed as phreatomagmatic or sometimes surtseyan tuff-rings, due to magma contacting water at shallow depths or in marginal lakes respectively. The Parade diatreme, Dunbar, contains over 300m of largely base-surge tuffs thought to represent the subsided inner flank deposits of a large maar. The Heads of Ayr and the East Fife diatremes expose different levels in subsided phreatomagmatic tuff-rings due to collapse-ahd erosion. Deep levels such as that exposed at Lundin Links, contain unbedded tuffs and abundant intrusive material. Shallower levels, such as at Elie Ness contain high proportions of bedded tuffs which are often centroclinally orientated. Base-surge, airfall, slumped and reworked tuffs in the Scottish diatremes are directly comparable to deposits in the modern tuff-rings studied proving their origin. A model for the formation of surtseyan tuff-rings is presented, with phreatomagmatic explosions resulting from steam expansion jets which disrupt an already vesiculating magma as it engulfes subsiding water-laden ash. A base surge model is also presented, involving deposition of tuffs with characteristic bedforms and structures by the head, body and tail of each surge analogous to turbidity currents. Cooling of hot, dry steam to cool moist steam towards the rear of surge pulses leads to lag breccias and progressive dune deposits being succeeded by regressive dunes and plastering structures with time. Juvenile sideromelane fragments erupted by phreatomagmatic volcanoes rapidly alter to palagonite as heated pore-waters circulate through the newly-deposited tuffs. Palagonitization results in cation mobility within unstable glass and precipitation of authigenic minerals in voids. Non-equilibrium growth of such minerals results in variable compositions and crystal forms. Subsequent alteration occurs slowly as a weathering process whose rate is greatly reduced as authigenic precipitation closes pore spaces within the tuffs. On diagenesis, unstable alteration products are commonly replaced by chlorite, calcite and clay. Reddening of some tuffs occurs by in situ breakdown of ironbearing minerals and release of Fe to solution although groundwater exchange with red country rock sediments may also occur. Unless present in diatremes phreatomagmatic products have a low preservation potential due to :- extreme alteration, rapid syn- and post-volcanic reworking, low ejecta volumes and breaching and burial beneath later lavas. In contrast the sedimentary structures petrography, morphology and grain size characteristics of diatreme tuffs are shown to be often sufficiently well preserved to permit the identification of their original surface volcanoes and their eruptive histories.

No description available.

Active lava lakes present a rare opportunity to observe directly the complex processes occurring within a magma body. Situated on Ross Island, Antarctica, the 3794-m-high crater of Erebus volcano has hosted a phonolite lava lake for decades. Previous studies have shown that many of the lake’s characteristics, such as surface velocity, gas flux and gas composition, exhibit a pronounced pulsatory behaviour on a time-scale of ∼10 min. Focusing primarily on the analysis of infra-red (IR) imagery acquired from the crater rim, this dissertation considers how the periodic behaviour of the Erebus lava lake evolves over decadal time periods, how the cyclic fluctuations of the different properties are interrelated and what can be inferred about the mechanisms occurring beneath the surface of the lake from these observations. Creation of new hardware, software and methodologies to facilitate these types of observations is a strong focus of this work. Chapter 1 introduces the nature of active lava lakes, reviews previous studies of Erebus and presents in detail the research objectives that are addressed by the subsequent chapters. In Chapter 2, a new thermal camera system that was developed as part of this study is described. Designed to run autonomously at the crater-rim of Erebus, this system was installed in December 2012 and has enabled, for the first time, extended time-series of images to be acquired. Chapter 3 briefly describes some of the other hardware and software that was developed as part of this study and outlines how it has been utilised for volcano monitoring. In Chapter 4, a dataset of IR images collected between 2004–2011 is used to assess inter-annual variability in the pulsatory behaviour of the surface motion of the Erebus lava lake. The cyclic behaviour is found to be a sustained feature of the lake, and no obvious changes are observed across the time period analysed. Data collected with the camera system described in Chapter 2 are analysed in Chapter 5 and combined with measurements from other instruments to assess the correlation between the cyclic behaviours of different lake properties. Cycles in surface speed, surface elevation, gas flux and gas composition are found to be highly correlated with each other. In Chapter 6, the surface velocities calculated in the preceding chapters are revisited, and the two-dimensional structure of the flow field is analysed. Chapter 7 demonstrates how the motion tracking methodologies developed for studying the Erebus lava lake can be used to improve high time resolution sulphur dioxide flux estimates - a significant challenge faced in the study presented in Chapter 5. Finally, Chapter 8 presents a synthesis of the key findings and conclusions from the preceding chapters.

No description available.

No description available.

No description available.