Caldera Rim Research Articles

Origin and geodynamic evolution of late Cenozoic potassium-rich volcanism in the Isparta area, southwestern Turkey

Post-collisional potassic-rich volcanism of Gölcük Volcano in the Isparta area of southwestern Turkey consists of two groups: (i) extracaldera lavas, corresponding mainly to Pliocene activity; and (ii) intracaldera lavas and pyroclastics (ignimbrite flows and ash/pumice fall deposits) formed during the Quaternary. Extracaldera volcanic rocks mainly comprise lamprophyre (minette), basaltic trachyandesite, trachyandesite, and trachyte. A close relationship exists between the silica content and phenocryst type in the extracaldera volcanics such that trachyte–trachyandesites with SiO2 < 57 wt% and basaltic trachyandesites are characterized mainly by mafic phenocryst phases (e.g. pyroxene, amphibole, biotite–phlogopite). These features suggest suppression of plagioclase crystallization under high H2O pressure conditions. Intracaldera volcanics are composed of tephriphonolitic dikes, remnants of lava flows and domes at the caldera rim, and a trachytic lava dome on the caldera floor. The Gölcük flows and pyroclastics are mainly characterized by strong incompatible element enrichment in large ion lithophile elements (LILEs; e.g. Cs, Ba, U, and Th) relative to K, Rb, and high-field strength elements (e.g. Nb, Ta, and Ti). We conclude that the Gölcük lavas were derived from a metasomatized lithospheric mantle source containing phlogopite–amphibole garnet peridotite; the latter resulted from metasomatism by a hydrous fluid phase related to subducted sediments and oceanic crust. The parental magma for the extracaldera volcanics was lamprophyric, and that for the intracaldera volcanics was basanitic. All the geological and geochemical data show that the alkaline Gölcük lavas display a gradual decrease in silica content with decreasing eruption age, indicating that, in the Isparta volcanic province, the asthenospheric melt component became more important over time. In the extracaldera volcanics, 87Sr/86Sr isotope ratios of the evolved trachyte–trachyandesites range between 0.70366 and 0.70504, whereas these ratios are lower in the less evolved basaltic trachyandesite and lamprophyres, varying in narrow ranges around 0.70365 and between 0.70374 and 0.70453, respectively. The 143Nd/144Nd values lie between 0.51264 and 0.51273 in trachyandesites, 0.51267 and 0.51273 in basaltic trachyandesites, and 0.51270 and 0.51274 in lamprophyres. In the intracaldera lavas, the 87Sr/86Sr isotope ratio is 0.70361 in the tephriphonolite and 0.70388 in the intracaldera trachytic lava dome. The 143Nd/144Nd isotope ratio is 0.51274 in the analysed tephriphonolitic flow and 0.51271 in the intracaldera trachytic lava dome; these values are higher than that of trachyte–trachyandesites of the extracaldera volcanics. The Sr–Pb isotopic signatures indicate that crustal contamination was significant for the evolved extracaldera volcanics, but was negligible for the intracaldera volcanics. The εNd values of the Gölcük volcanics range between 0 and 2.0. The low Sr isotope ratios and positive εNd values are characteristic features of a depleted mantle source. The isotopically depleted and incompatible enriched nature of the Gölcük lavas point to recent enrichment processes prior to partial melting of the mantle source. Conversely, their radiogenic lead isotope compositions (206Pb/204Pb = 19.19–19.54, 207Pb/204Pb = 15.64–15.67, 208Pb/204Pb = 39.12–39.49) indicate an enriched mantle source region.

The Incapillo Caldera and Dome Complex (∼ 28° S, Central Andes): A stranded magma chamber over a dying arc

The Plio-Pleistocene Incapillo Caldera and Dome Complex (5570 m) is situated at the southernmost termination of the Andean Central Volcanic Zone (∼ 28° S), where the more steeply dipping segment of the subducting Nazca plate transitions into the Chilean–Pampean “flat-slab” to the south. The ∼ 250 m high walls of the elliptical caldera (6 × 5 km) are composed of a homogeneous pumice-rich rhyodacitic ignimbrite that radiates ∼ 15 km from the caldera rim. New biotite 40Ar/ 39Ar ages along with previous K–Ar ages reveal that the final eruption of the Incapillo Caldera occurred at ∼ 0.51 ± 0.04 Ma. Flanking the caldera are ∼ 40 steep-sided rhyodacitic domes that predate the main “caldera-forming” ignimbrite. The morphology of these domes reflects the non-explosive effusive eruption of a crystal-rich gas-poor rhyodacitic magma. A volumetrically-minor debris flow composed of abundant pumice, dacites, and lithic clasts is topographically confined to a glacial valley south of the caldera and unconformably overlies the ignimbrite. Mixed pumices at the distal end of the ignimbrite reflect the convective physical condition within the magma chamber immediately prior to the pyroclastic eruption. Including the domes and ignimbrite, the total estimated volume of erupted Incapillo magma is ∼ 37 km 3. Based on its age, small size, and effusive eruptive style, the Incapillo Caldera and Dome Complex provides an endmember model for an Andean caldera erupting within a waning magmatic arc over a shallowing subduction zone.

A 90,000-year tephrostratigraphic framework of Aso Volcano, Japan

A detailed 90,000-year tephrostratigraphic framework of Aso Volcano, southwestern Japan, has been constructed to understand the post-caldera eruptive history of the volcano. Post-caldera central cones were initiated soon after the last caldera-forming pyroclastic-flow eruption (90 ka), and have produced voluminous tephra and lava flows. The tephrostratigraphic sequence preserved above the caldera-forming stage deposits reaches a total thickness of 100 m near the eastern caldera rim. The sequence is composed mainly of mafic scoria-fall and ash-fall deposits but 36 silicic pumice-fall deposits are very useful key beds for correlation of the stratigraphic sequence. Explosive, silicic pumice-fall deposits that fell far beyond the caldera have occurred at intervals of about 2500 years in the post-caldera activity. Three pumice-fall deposits could be correlated with lava flows or an edifice in the western part of the central cones, although the other silicic tephra beds were erupted at unknown vents, which are probably buried by the younger products from the present central cones. Most of silicic eruptions produced deposits smaller than 0.1 km 3, but bulk volumes of two silicic eruptions producing the Nojiri pumice (84 ka) and Kusasenrigahama pumice (Kpfa; 30 ka) were on the order of 1 km 3 (VEI 5). The largest pyroclastic eruption occurred at the Kusasenrigahama crater about 30 ka. This catastrophic eruption began with a dacitic lava flow and thereafter produced Kpfa (2.2 km 3). Total tephra volume in the past 90,000 years is estimated at about 18.1 km 3 (dense rock equivalent: DRE), whereas total volume for edifices of the post-caldera central cones is calculated at about 112 km 3, which is six times greater than the former. Therefore, the average magma discharge rate during the post-caldera stage of Aso Volcano is estimated at about 1.5 km 3/ky, which is similar to the rates of other Quaternary volcanoes in Japan.

The Gravity Anomaly of Ungaran Volcano, Indonesia : Analysis and Interpretation

Ungaran Volcano is located in the province of Central Java, Indonesia, and is a Quaternary volcano that consists of older and younger volcanoes. The older Ungaran volcano formed over 500,000 years ago, and the younger volcano was active until 300,000 years ago. The younger volcano seems to have been constructed inside a caldera formed by the older Ungaran activity. In this study, gravity data was used in an attempt to determine the exact location of the younger and older Ungaran volcanoes, and to investigate the relationship between fault structure and geothermal manifestations. A positive Bouguer anomaly was observed over the volcanic body. From detailed analysis of the gravity data, high anomalies were located over the northern part of the summit that correlate with the older Ungaran volcano. Various interpretation methods, such as horizontal gradient analysis, spectral analysis, and 2-D forward modeling, were applied to the gravity data. The younger Quaternary volcanic rocks, which consist of hornblende-augite-andesite (andesite lava) of the Gajahmungkur volcanics, have an average density of 2,390 ± 120 kg/m 3 , while the older Quaternary volcanic rocks, consisting of augite-olivine basalt flows (basaltic lava) from the Kaligesik formation, have an average density of 2,640 ± 100 kg/m 3 . The structural setting of the Ungaran volcano has characterized by circular structure where most geothermal manifestations are located. The result of gravity analysis shows that Ungaran volcano seems to have occurred in tectonic depression, and prominent caldera depression has not formed within Ungaran volcano without surficial caldera rim. The horizontal gradient analysis indicates that geothermal features at Ungaran volcano are structurally controlled and are located within the younger volcano.

Asymmetric growth of collapsed caldera by oblique subsidence during the 2000 eruption of Miyakejima, Japan

Oblique development of the ring faults reflecting the structural heterogeneities inside the volcano formed many asymmetric structures of Miyakejima 2000 AD caldera. The asymmetry includes (a) offset location of the ring faults with respect to the associated shallow magma chamber, (b) unequal outward migration of the caldera wall 600 m at the southeastern rim but only 200 m at the northwestern rim, (c) development of tilted terrace only at the southeastern caldera margin, (d) eruption sites and fumaroles being confined to the southern part of the caldera. Geophysical data, including ground deformation and seismic activity, indicates the offset of the location of the magma chamber about 2 km south of the caldera center on the surface. The ring faults propagated from the deflating magma chamber obliquely about 30 degrees toward the summit. The oblique subsidence of the cylindrical block formed a wider instable zone, particularly in the southeastern side of the ring fault that enhanced the larger outward migration of the caldera rim and also caused the formation of the outer half-ring fault bordering the tilting slope at the southern part. Ascending pass of the buoyant magma along the tilted ring faults was concentrated in the southern half of the caldera and consequently the distributions of the eruption sites and fumaroles are localized in the southern-half part of the caldera. The structure of the Miyakejima 2000 caldera with complete development of the ring faults, its high roof aspect ratio and oblique subsidence is clearly distinguishable from trapdoor-type caldera. The oblique development of the ring faults can be controlled by the mechanical contrast between the solidified conduits and surrounding fragile volcanic edifice. Asymmetric development of the Miyakejima caldera shows that the collapsed calderas are potential indicators of the heterogeneous structures inside of the volcano, particularly in the case of small-size caldera.

Structure and CO2 budget of Merapi volcano during inter-eruptive periods

Soil temperature and gas (CO2 concentration and flux) have been investigated at Merapi volcano (Indonesia) during two inter-eruptive periods (2002 and 2007). Precise imaging of the summit crater and the spatial pattern of diffuse degassing along a gas traverse on the southern slope are interpreted in terms of summit structure and major caldera organization. The summit area is characterized by decreasing CO2 concentrations with distance from the 1932 crater rim, down to atmospheric levels at the base of the terminal cone. Similar patterns are measured on any transect down the slopes of the cone. The spatial distribution of soil gas anomalies suggests that soil degassing is controlled by structures identified as concentric historical caldera rims (1932, 1872, and 1768), which have undergone severe hydrothermal self-sealing processes that dramatically lower the permeability and porosity of soils. Temperature and CO2 flux measurements in soils near the dome display heterogeneous distributions which are consistent with a fracture network identified by previous geophysical studies. These data support the idea that the summit is made of isolated and mobile blocks, whose boundaries are either sealed by depositional processes or used as pathways for significant soil degassing. Within this context, self-sealing both prevents long-distance soil degassing and controls heat and volatile transfers near the dome. A rough estimate of the CO2 output through soils near the dome is 200–230 t day−1, i.e. 50% of the estimated total gas output from the volcano summit during these quiescent periods. On Merapi’s southern slope, a 2,500 m long CO2 traverse shows low-amplitude anomalies that fit well with a recently observed electromagnetic anomaly, consistent with a faulted structure related to an ancient avalanche caldera rim. Sub-surface soil permeability is the key parameter that controls the transfer of heat and volatiles within the volcano, allowing its major tectonic architecture to be revealed by soil gas and soil temperature surveys.

Gravity‐driven deformation of Tenerife measured by InSAR time series analysis

We study the state of deformation of Tenerife (Canary Islands) using Differential Synthetic Aperture Radar Interferometry (DInSAR). We apply the Small BAseline Subset (SBAS) DInSAR algorithm to radar images acquired from 1992 to 2005 by the ERS sensors to determine the deformation rate distribution and the time series for the coherent pixels identified in the island. Our analysis reveals that the summit area of the volcanic edifice is characterized by a rather continuous subsidence extending well beyond Las Cañadas caldera rim and corresponding to the dense core of the island. These results, coupled with GPS ones, structural and geological information and deformation modeling, suggest an interpretation based on the gravitational sinking of the dense core of the island into a weak lithosphere and that the volcanic edifice is in a state of compression. We also detect more localized deformation patterns correlated with water table changes and variations in the deformation time series associated with the seismic crisis in 2004.

Converted phases analysis of the Campi Flegrei caldera using active seismic data

We applied a type of depth migration for converted seismic phases to active and passive seismic data sets from the Campi Flegrei volcanic region in southern Italy. The migration method is based on the diffraction summation migration technique. Travel times to grid points in a volume were calculated in smooth P and S-velocity models and trace energy near the calculated converted phase time was stacked over multiple sources at one receiver. Weighting factors based on Snell's Law at an arbitrarily oriented local interface were applied to better focus trace energy. PP reflection images from the active data set provide increased detail to images of the caldera rim from other studies. The migrated images also show features near 2–3 km in depth beneath Pozzuoli City, which may be associated with an over-pressured gas volume, as suggested by other geophysical investigations. Possible deeper features near 4 km depth may be related to the presence of the carbonate basement or may image a previously undetected feature, such as a small body of strongly thermometamorphosed volcanic rock. The current passive earthquake data set from the 1984 ground uplift episode was not well suited to the converted phases analysis due to narrow P–S windows and high noise levels in the traces. However, two stations provide confirmation and extension of imaged features in the active data.

How summit calderas collapse on basaltic volcanoes: New insights from the April 2007 caldera collapse of Piton de la Fournaise volcano

In April 2007, Piton de la Fournaise volcano experienced a caldera collapse during its largest historical eruption. We present here a structural analysis both of the caldera and the surrounding area, and precise GPS data recorded with a dense GPS network specifically dedicated to the analysis of deformation related to the summit collapse structures. Despite a collapse of more than 300 m in the central zone, the geometry of the new caldera is similar in map view to that of the pre-existing collapsed structure, which was formed from the coalescence of several pit craters. The caldera shows an asymmetric inner geometry with sub-vertical walls in the NW quadrant and steep scarps composed of inward tilted blocks in the southern half. The presence of preserved polished surfaces on the lower part of the sub-vertical scarp indicates that it corresponds to the caldera north-western ring fault. The April 2007 caldera collapse led to the development and the reactivation of concentric fractures on the caldera rim, mostly along the southern limit of the caldera. GPS data show that fractures result from radial extensional stresses that are restricted within the first tens of meters of the caldera edge. GPS data also reveal that the caldera collapse was coeval with a centripetal deflation, whose magnitude is largest along the southern half of the caldera. The displacements recorded by GPS result from both a general deflation, due to magma withdrawal from Piton de la Fournaise's summit magma chamber, and additional local effects related to the caldera collapse. Comparison of the caldera collapses at Piton de la Fournaise, Miyakejima and Fernandina reveals striking similarities, with cyclic seismic signals accompanying small-scale deflation–inflation cycles. This strongly suggests a common mode of collapse. Hence, we propose a unifying model of caldera collapse in basaltic setting, in which the inward deflation due to magma withdrawal from the magma chamber prevents the collapse of the caldera roof until the gravitational stress acting on the rock column above the magma chamber exceeds the shear strength along pre-existing ring faults. The downward displacement stops when the pressure increase into the magma chamber is able to again sustain the rock column. The succession of (1) inward deflation that prevents the collapse, (2) collapse due to gravitational stress and (3) stopping of the downward motion is repeated many times. The frequency of the cycles is influenced by the rate of magma withdrawal and by the amount of intrusion of magma along the ring faults.

Geologic evolution of the Donguinyó-Huichapan caldera complex, central Mexican Volcanic Belt, Mexico

The Donguinyó-Huichapan caldera complex is located 110 km to the NNW of Mexico City, in the central sector of the Mexican Volcanic Belt. It is a 10 km in diameter complex apparently with two overlapping calderas, each one related to an ignimbrite sequence that contrasts in composition, mineralogy, welding, distribution, and physical aspect. The geologic evolution of this complex includes the following phases, 1) A first caldera formed at 5.0 ± 0.3 Ma, with the eruption of several discrete pulses of andesitic to trachydacitic pyroclastic flows that produced a series of densely welded ignimbrites; 2) At 4.6 ± 0.3 Ma, several small shield volcanoes and cinder cones built the rim of this caldera and erupted basaltic-andesite and andesitic lava flows; 3) At 4.2 ± 0.2 Ma, a second caldera was formed associated to the eruption of the Huichapan Tuff, which is a rhyolitic pyroclastic sequence consisting of minor unwelded ignimbrites, pumice fall and surge deposits, and a voluminous welded ignimbrite; 4) Also yielding an age of 4.2 ± 0.2 Ma, several trachydacitic lava domes were extruded along the new ring fracture and formed the rim of the Huichapan caldera, as well as five intra-caldera domes of dacitic and trachydacitic composition. Peripheral volcanism includes a large 2.5 ± 0.1 Ma shield volcano that was emplaced on the Huichapan caldera rim. The two calderas that form the Donguinyó-Huichapan complex have contrasting differences in volcanic styles that were apparently due to their differences in composition. Products erupted by the Donguinyó caldera are basaltic-andesite to trachydacitic in composition, whereas Huichapan caldera products are all high-silica rhyolites.

Airborne laser swath mapping of the summit of Erebus volcano, Antarctica: Applications to geological mapping of a volcano

The lidar observations reported here have provided the first high precision topographic data on the active craters of Erebus volcano, Antarctica. Digital elevation data are invaluable for geomorphic analysis of volcanoes especially when combined with methods to automatically extract landscape features. The data are indispensable for studying eruptive events and volcano-tectonic processes. Airborne laser scanning or lidar is a rapidly emerging technology to map surface topography with high accuracy and detail, comparable to a 1:1000 scale map. In the austral summer of 2001–2002 airborne laser scanning data were collected in Antarctica over the summit plateau of Erebus volcano. We describe the high-resolution data set from Erebus, products derived from the data and demonstrate how extracted topographic features can be used in the geomorphic analysis of volcanoes. Background is presented on the airborne laser scanning method and on the accuracy estimation of laser points. We describe the computation of a high-resolution DEM with a grid size of 2 m × 2 m for the summit area of Erebus volcano. Errors in the DEM are assessed using independent ground-based GPS data and the RMS error is estimated to be ± 0.39 m. Visualization methods reveal topographic features that are very useful for geomorphic analysis. Automatic feature extraction and processing is necessary to exploit the high-resolution and accurate elevation data. Methods are described to extract 1D and 2D features and organize them for subsequent analysis. Fitting analytical functions to 3D point clouds or DEMs is particularly effective in extracting topographic features, including planes, higher order surfaces, cones, and conic sections. These methods are used to obtain a quantitative description of the calderas and modern-day crater of Erebus. Especially valuable are the measured surface areas if the lava lakes as these control thermal emissions and gas fluxes from the volcano. Examination of the surface elevations of Ray Lava Lake, Werner Vent and Active Vent indicates that the active vents were more or less in magmastatic equilibrium in 2001 December. Conic segments are fitted to mapped sections of caldera rims to identify the sections buried under younger lava flows or covered by ice and snow. Automatic methods are used to detect and characterize lava channels using geomorphometric maps generated from the DEM.

Refinement of the late Quaternary geologic history of Erebus volcano, Antarctica using 40Ar/ 39Ar and 36Cl age determinations

Six new 40Ar/ 39Ar and three cosmogenic 36Cl age determinations provide new insight into the late Quaternary eruptive history of Erebus volcano. Anorthoclase from 3 lava flows on the caldera rim have 40Ar/ 39Ar ages of 23 ± 12, 81 ± 3 and 172 ± 10 ka (all uncertainties 2 σ). The ages confirm the presence of a second, younger, superimposed caldera near the southwestern margin of the summit plateau and show that eruptive activity has occurred in the summit region for 77 ± 13 ka longer than previously thought. Trachyte from “Ice Station” on the eastern flank is 159 ± 2 ka, similar in age to those at Bomb Peak and Aurora Cliffs. The widespread occurrences of trachyte on the eastern flank of Erebus suggest a major previously unrecognized episode of trachytic volcanism. The trachyte lavas are chemically and isotopically distinct from alkaline lavas erupted contemporaneously in the summit region < 5 km away. The three 36Cl ages are the first exposure ages reported for rocks from Erebus volcano. At an assumed erosion rate of 1.5 mm/ka 36Cl cosmogenic ages are 6.0 ± 0.5 ka for the Ice Tower Ridge flow, 4.9 ± 0.4 ka for Lower Hut flow, and 6.5 ± 0.6 ka for the Northeast flow. The 36Cl ages overlap within analytical uncertainty or are slightly younger than 40Ar/ 39Ar ages for these lava flows, and confirm the overall young age (< 25 ka) of the summit lava flows. One interpretation of these uniformly young ages is that older eruptions infilled the topographic depression caused by caldera formation and subsequently was covered by younger lavas.

Caldera rim collapse: A hidden volcanic hazard

Following the emblematic flank collapse of Mount St Helens in 1981, numerous models of flank sliding have been proposed. These models have allowed to largely improve the understanding of mechanisms involved in such landslides, which represent a tremendous risk for populations living around volcanoes. In this article, a new mode of landslide formation, related to buried calderas, is described. The model emphasizes the paramount importance of the hidden ring fault that, even when the caldera is buried, still remains a plane of weakness in the core of the edifice. Under certain conditions, this plane of weakness becomes activated as the upper part of a pre-existing critical slip surface and is used in the emplacement of huge landslides which travel downslope at a very high velocity. A natural example is taken from Piton de la Fournaise Volcano (La Réunion Island, Indian Ocean). It reveals that the primary cause triggering caldera rim collapse is partial unbuttressing of the flank of the volcano. In the natural example, this occurs through regressive erosion that excavates deep canyon in the direction of the buried caldera but other mechanisms may exist. On account of the large volumes of material involved in caldera rim collapse as well as their long runout distances, such a volcanic hazard should be taken into account on every volcano where buried calderas are suspected.

Kaguyak dome field and its Holocene caldera, Alaska Peninsula

Kaguyak Caldera lies in a remote corner of Katmai National Park, 375 km SW of Anchorage, Alaska. The 2.5-by-3-km caldera collapsed ~ 5.8 ± 0.2 ka ( 14C age) during emplacement of a radial apron of poorly pumiceous crystal-rich dacitic pyroclastic flows (61–67% SiO 2). Proximal pumice-fall deposits are thin and sparsely preserved, but an oxidized coignimbrite ash is found as far as the Valley of Ten Thousand Smokes, 80 km southwest. Postcaldera events include filling the 150-m-deep caldera lake, emplacement of two intracaldera domes (61.5–64.5% SiO 2), and phreatic ejection of lakefloor sediments onto the caldera rim. CO 2 and H 2S bubble up through the lake, weakly but widely. Geochemical analyses ( n = 148), including pre-and post-caldera lavas (53–74% SiO 2), define one of the lowest-K arc suites in Alaska. The precaldera edifice was not a stratocone but was, instead, nine contiguous but discrete clusters of lava domes, themselves stacks of rhyolite to basalt exogenous lobes and flows. Four extracaldera clusters are mid-to-late Pleistocene, but the other five are younger than 60 ka, were truncated by the collapse, and now make up the steep inner walls. The climactic ignimbrite was preceded by ~ 200 years by radial emplacement of a 100-m-thick sheet of block-rich glassy lava breccia (62–65.5% SiO 2). Filling the notches between the truncated dome clusters, the breccia now makes up three segments of the steep caldera wall, which beheads gullies incised into the breccia deposit prior to caldera formation. They were probably shed by a large lava dome extruding where the lake is today.

Eruptive history and tectonic setting of Medicine Lake Volcano, a large rear-arc volcano in the southern Cascades

Medicine Lake Volcano (MLV), located in the southern Cascades ∼ 55 km east-northeast of contemporaneous Mount Shasta, has been found by exploratory geothermal drilling to have a surprisingly silicic core mantled by mafic lavas. This unexpected result is very different from the long-held view derived from previous mapping of exposed geology that MLV is a dominantly basaltic shield volcano. Detailed mapping shows that < 6% of the ∼ 2000 km 2 of mapped MLV lavas on this southern Cascade Range shield-shaped edifice are rhyolitic and dacitic, but drill holes on the edifice penetrated more than 30% silicic lava. Argon dating yields ages in the range ∼ 475 to 300 ka for early rhyolites. Dates on the stratigraphically lowest mafic lavas at MLV fall into this time frame as well, indicating that volcanism at MLV began about half a million years ago. Mafic compositions apparently did not dominate until ∼ 300 ka. Rhyolite eruptions were scarce post-300 ka until late Holocene time. However, a dacite episode at ∼ 200 to ∼ 180 ka included the volcano's only ash-flow tuff, which was erupted from within the summit caldera. At ∼ 100 ka, compositionally distinctive high-Na andesite and minor dacite built most of the present caldera rim. Eruption of these lavas was followed soon after by several large basalt flows, such that the combined area covered by eruptions between 100 ka and postglacial time amounts to nearly two-thirds of the volcano's area. Postglacial eruptive activity was strongly episodic and also covered a disproportionate amount of area. The volcano has erupted 9 times in the past 5200 years, one of the highest rates of late Holocene eruptive activity in the Cascades. Estimated volume of MLV is ∼ 600 km 3, giving an overall effusion rate of ∼ 1.2 km 3 per thousand years, although the rate for the past 100 kyr may be only half that. During much of the volcano's history, both dry HAOT (high-alumina olivine tholeiite) and hydrous calcalkaline basalts erupted together in close temporal and spatial proximity. Petrologic studies indicate that the HAOT magmas were derived by dry melting of spinel peridotite mantle near the crust mantle boundary. Subduction-derived H 2O-rich fluids played an important role in the generation of calcalkaline magmas. Petrology, geochemistry and proximity indicate that MLV is part of the Cascades magmatic arc and not a Basin and Range volcano, although Basin and Range extension impinges on the volcano and strongly influences its eruptive style. MLV may be analogous to Mount Adams in southern Washington, but not, as sometimes proposed, to the older distributed back-arc Simcoe Mountains volcanic field.

Volcanic features and the hydrological setting of Southern Thule, South Sandwich Islands

AbstractThis paper provides new observations of volcanic features and hydrological characteristics in and around Southern Thule, the southernmost group of islands in the South Sandwich Islands, including the first high-resolution bathymetric image of the Douglas Strait caldera. The South Sandwich Islands are the summits of several very large subduction-related volcanoes constructed at the eastern boundary of the Scotia Sea. Observations of the islands are scarce owing to their remote location and they are only rarely visited, yet the area is an active volcanic arc that is rapidly changing as a result of eruptions, including one (on Montagu Island) that has been ongoing for six years and is creating new land. The three islands that make up Southern Thule are morphologically different, and they illustrate different stages in the construction and evolution of islands in the South Sandwich group. We present the results of an acoustic and hydrographic survey that resulted in the first high-resolution, multibeam ‘swath’ image of the submarine Douglas Strait caldera. The results confirm the presence of a large sediment mound (c. 1000 m3 in volume) on the floor of the Douglas Strait caldera related to a flank collapse of Thule Island. The image also shows an extensive arcuate fault structure, interpreted as evidence that the caldera is nested, and its geomorphological freshness suggests that it is a very young feature that formed conceivably in the last few decades or centuries. The bathymetric image also reveals at least three cone- or mound-like structures within the caldera that may relate to renewed post-caldera volcanism (as pyroclastic cones and/or pillow mounds). Recently formed cones and faults associated with caldera rims are often associated with hydrothermal activity. However, temperature and salinity data collected within the caldera do not yet show any evidence of hydrothermal venting. The ocean surrounding Southern Thule has a prominent surface layer of low salinity water that is probably caused by a high meltwater flux derived from ice caps on the islands. This flux may have been enhanced by the recent warming trend observed in the Antarctic Peninsula, although volcano-related geothermal melting in the ice-filled Thule Island caldera may also contribute to the meltwater flux.

Quilotoa volcano — Ecuador: An overview of young dacitic volcanism in a lake-filled caldera

Quilotoa volcano, an example of young dacitic volcanism in a lake-filled caldera, is found at the southwest end of the Ecuador's volcanic front. It has had a long series of powerful plinian eruptions of moderate to large size (VEI = 4–6), at repetitive intervals of roughly 10–15 thousand years. At least eight eruptive cycles (labeled Q-I to Q-VIII with increasing age) over the past 200 ka are recognized, often beginning with a phreatomagmatic onset and followed by a pumice-rich lapilli fall, and then a sequence of pumice, crystal, and lithic-rich deposits belonging to surges and ash flows. These unwelded pyroclastic flows left veneers on hillsides as well as very thick accumulations in the surrounding valleys, the farthest ash flow having traveled about 17 km down the Toachi valley. The bulk volumes of the youngest flow deposits are on the order of 5 km 3, but that of Q-I's 800 yr BP ash-fall unit is about 18 km 3. In the last two eruption cycles water has had a more important role. Upon Quilotoa's low-relief volcanic edifice, three calderas are recognized; the formation of the oldest one predates the Q-IV cycle and the others occurred during the Q-II and Q-I cycles. Dacite lava domes are common along the present caldera rim and most were emplaced at the end of the Q-II cycle; older domes of dark dacite belong to the Q-III and IV cycles. The explosive onset of the Q-I cycle expulsed as much as 250 million m 3 of the lake's water, resulting in large debris flows that scoured the eastern flanks of the edifice and descended the Toachi river. Little variation in the mineralogy and chemistry of Quilotoa's eruptive products is observed, suggesting that the source is a homogeneous magma body at shallow depth. Both the pyroclastic material and the domes are composed chiefly of gray porphyritic dacites carrying large phenocrysts of plagioclase, amphibole, biotite, and occasionally quartz.

New K–Ar age determinations of Kilimanjaro volcano in the North Tanzanian diverging rift, East Africa

The Kilimanjaro is the African highest mountain and culminates at 5895 m high. This huge volcanic edifice is composed of three main centres along a N110°E-striking axis (Shira, Kibo and Mawenzi from W to E), and emplaced in a key area where a major N80°E-oriented volcanic lineament intersects a first-order NW–SE basement fault-like discontinuity. Seventeen K–Ar ages (on microcrystalline groundmass) acquired on lavas and intrusive facies from the three eruptive centres confirm that the Plio-Quaternary volcanicity of Kilimanjaro is clearly polyphased. The oldest phases of volcanic activity begun at ~ 2.5 Ma in the Shira vent and our data suggest that the latest important phases occurred around 1.9 Ma, just before the collapse of the Northern part of the edifice. Magmatic activity then shifted eastwards in the Mawenzi and Kibo twin centres where initial volcanism is dated at ~ 1 Ma. Two K–Ar ages obtained for the most recent Mawenzi rocks from the Neumann Tower-Mawenzi group (492 ka) and Mawenzi eruptive centre (448 ka), near the present summit, are linked to the final stage of edification for this centre. Whereas the eruptive activity ceased in the Mawenzi, it still continued on Kibo since sub-actual time. The oldest dated rocks from Kibo (482 ka) is obtained on a dyke from the Lava Tower group cropping out at 4600 m high. The main phase of magmatism on Kibo is recorded by two lava formations with a great spatial extension – the Rhomb Porphyry group and the Lent group – that have been emplaced in a short time interval at ~ 460–360 ka (including two erosive stages) and 359–337 ka, respectively. Based on the dating of Caldera rim group lavas, it is shown that the edification of the present cone was accomplished in a period ranging from 274 to 170 ka. The new ages obtained for the main episodes of volcanic activity on Kibo appear to roughly coincide with the oldest known Quaternary glaciations. The interaction between eruptive phenomena and the ice cover is assumed to have played an important role in triggering collapse processes and associated lahars deposits. The last volcanicity, around 200–150 ka, is marked by the formation of the present summit crater in Kibo and the development of linear parasitic volcanic belts, constituted by numerous Strombolian-type isolated cones on the NW and SE slopes of Kilimanjaro. These belts are likely to occur above deep-seated fractures that have guided the magma ascent, and the changes in their directions with time might be related to the rotation of recent local stress field.

Contrasting styles of welding observed in the proximal Askja 1875 eruption deposits I: Regional welding

Welded fall deposits on the northern caldera rim at Askja volcano are associated with the Plinian phase of the 1875 eruption. Two welding units occur within the proximal Plinian fall centered on stratigraphic sub-units which, where non-welded, are poorly sorted and ash-rich with high abundances of fluidal and needle-like ash particles. Welding has formed due to two discrete processes; a) the sintering of hot ash and lapilli which forms the two distinct units that are laterally continuous on distance scales of tens of meters (termed ‘regional welding’), and b) creation of welding halos enclosing large, dense, discrete, non- to poorly vesicular spatter bombs that are up to 9 m in diameter (termed ‘local welding’). This paper is concerned with the nature of regional welding and the companion paper (this issue) focuses on the phenomenon of local welding. Three case studies documenting the range of welding patterns observed in regional welding are presented here. Vertical and lateral profiles of welding intensity, together with the deposit characteristics reveal that welding could only occur when the accumulation rates were sufficient and that grain size and thickness are second order factors facilitating welding. Rapid and reversible shifts in both thickness and welding grade are observed on a ~ 10 m scale laterally along the caldera rim suggesting considerable unsteadiness of the transport regime, which promoted localized fluctuations of the accumulation rate. The welded deposits prompt re-examination of both the dynamics of the Plinian phase of the 1875 eruption and the distribution of source vents. The dispersal of the welding units is not compatible with deposition from the full height of the Plinian plume. Similarly to the ultra-proximal deposits of Novarupta or Tarawera, these clasts probably fell from heights of hundreds of meters to < 4 km, retaining sufficient heat to weld after deposition. The E–W elongated distribution of the welded units is also not compatible with a single source vent, and favors several vents that were in a fountaining phase, and located adjacent to the northern rim.

Contrasting styles of welding observed in the proximal Askja 1875 eruption deposits II: Local welding

As an alternative to classical welding models of fall deposits due to the progressive accumulation of hot tephra which then weld, we describe here welded deposits on the northern 1875 caldera rim of Askja volcano that have welded due to the influence of hot, discrete spatter bombs impacting into and supplying heat to a halo of surrounding tephra. This style of welding we term ‘local welding’ in contrast to ‘regional welding’ which is described elsewhere [Carey, R.J., Houghton, B.F., Thordarson, T., 2008. Contrasting styles of welding observed in the proximal Askja 1875 eruption deposits I: Regional welding. J. Volcanol. Geotherm. Res. 171, 1–19. doi:10.1016/j.jvolgeores.2007.11.020]. Locally welded deposits are associated with the rhyolitic Plinian phase of the 1875 eruption of Askja volcano. Two distinct welding units (W1 and W2) are interbedded with Plinian fall on the northern caldera rim, and grade outwards to weakly dispersed non-welded fall. Spatter bombs are found in both welding units but vary in their characteristic sizes and internal features. In the W1 unit simple bombs with homogeneous internal characteristics up to ∼ 60 cm in diameter are found. In the W2 unit, large discrete spatter bombs with complex internal features range up to 9 m in diameter. We describe here two case studies showing the effects of a) single small spatter bombs; b) multiple small spatter bombs and c) large discrete spatter bombs varying in size. Vertical and lateral profiles through welding zones reveal that the primary controls on local welding are the availability of supplied or added heat and the loading capacity of the spatter bomb. Local welding grades are much higher than that of regional welding, as the combined effects of heat, compaction and insulation can provide suitable conditions which lead to dense welding and, proximal to the spatter bomb, rheomorphic flowage. If heating and loading exceed the critical requirement for welding, porosity loss via matrix welding and vesicle collapse occurs to a point where further strain must be accommodated as shearing and ductile flowage. The spatter bombs are found only within the weakly dispersed welding units and are the final erupted products of each fountaining phase. Their low viscosities are evident by their deformation on impact and fluidal forms, and hold some important clues to eruption dynamics in the shallow conduit and vent regions.