Iceland is the largest portion of the mid-Atlantic ridge exposed above sea level. It is the result of the interplay between a mantle plume and the mid-oceanic ridge resulting in a localised concentration of volcanic activity. As a result Iceland is the best place to study plume - mid ocean ridge interaction, crustal deformation at divergent plate boundary, extensional volcanism and formation of the ocean crust during rifting. Due to its location in the north Atlantic, Iceland has experienced numerous glaciations and is a place of choice to study ice-magma interaction, crustal rebound and its effect on volcanism and glacier erosion and depositional processes. This report is a wide-ranging synopsis of the geology of Iceland and is based on a fourteen day field trip around the Island. Although a variety of subjects are exposed, the focus is on volcanism. The main objective of this trip was to recognise volcanic textures and structures and understand their origins and implications in term of volcanic processes as well as to gain a sense of the uniqueness of Iceland as a geo-laboratory. This report is divided in four chapters; Tectonic setting, Geological framework, Volcanism, and Geothermal activity.
Iceland is located in
the North Atlantic Large Igneous Province (NALIP) which erupted during the
opening of the North Atlantic 55-62 Ma ago. Iceland is part of the second phase
of igneous activity which begins about 56 Ma ago (Saunder et al, 1997) producing
up to 10 million km3 of igneous rock during 2-3 million years (White
and McKenzie, 1989). Volcanic material from that period are found on both sides
of the north Atlantic and include the onshore Tertiary igneous provinces of
Britain, Northern Ireland, the Faeroes, Greenland and Baffin Island (phase 1) as
well as extensive submarine volcanic rifted margins (figure 2). Iceland is centered on the mid-Atlantic ridge,
diffracting it between the Reykjanes Ridge to the south and the Kolbeinsey
ridge to the north. In this chapter we explore the effect of the ridge and
spreading center in Iceland; next we look at the probable mantle plume
underneath the island and then consider the interplay of the two.
The Mid-Atlantic ridge is composed of a series of spreading centers that are offset by transform faults which concentrate the seismic activity of the extension zone. South of Iceland the Reykjanes Ridge comes onshore at the southern tip of the Reykjanes Peninsula and its surface manifestation can be observed on land at the location of the first Icelandic parliament (AlÞing in 930) at the Unesco site of Þingvellir (figure 3). This site displays the typical morphology of a lithosphere under tension forming a graben system with isolated horst and normal motion on fault plans. From a structural point of view, continental extension has been subdivided into narrow rifting, wide rifting and core complex modes (Buck, 1991). Iceland qualifies as a narrow rifting zone and is closer to a typical oceanic rift zone than any other continental rift. Narrow rifts are characterized by a concentrated crustal and mantle lithospheric extension that gives rise to narrow regions (generally up to 100–150 km wide but only 30 km wide in Iceland) of intense normal faulting (Buck, 1991). Unlike core complex type extension, narrow rift zones are characterized by high angle normal fault arranged in a two sided system dipping inward as discernible in Þingvellir. Stretching of the continental lithosphere develops in response to regional stress field (passive rifting) or thermal upwelling of the asthenosphere (active rifting) (Cortis et al, 2003). In the case of Iceland active rifting is the dominant process with an ascending mantle plume added to an oceanic ridge system causing lithosphere thinning and isostatic crustal doming. In the next section we will look at evidences for a mantle plume beneath Iceland.
The Iceland hotspot is commonly thought to be the surface representation of a mantle plume. This is the explanation of choice for the gravity and topography data rising well above the rest of the mid-Atlantic ridge. The term Plume has many definitions and can be viewed as a convecting upwelling feature that is maintained by thermal buoyancy. It is thought to originate from a thermal boundary layer such as the 660 km boundary (between the upper and lower mantle marking the phase change from Ringwoodite to Perovskite), or the 2700-2900km boundary (the D’’ boundary between lower mantle and outer core, with the appearance of a post-perovskite phase at 2700km (Hirose, 2007)). High-pressure experimental studies of the melting point of iron–nickel alloys show that the core is several hundred degrees hotter than the overlying mantle. A temperature difference of this magnitude is expected to produce an unstable boundary layer above the core which, in turn, should produce plumes of hot, solid material that rise through the mantle, driven by their thermal buoyancy. Therefore, from theoretical considerations, mantle plumes are the inevitable consequence of a hot core and a sulfate melt /cooler silicate melt contact (Campbell, 2005).
Several attempts to image the Icelandic plume by seismic experiments provide similar conclusions for the first hundreds of kilometers but all suffer from a resolution problem under a depth of 400 km. Since the 6o’s the presence of an anomalous low seismic velocity zone under Iceland is known. Recent mapping of this zone by Allen et al (2002) using body and surface wave data show a cylindrical low velocity anomaly extending from the surface to a depth of 400km (Figure 4a). The resolution problem at lower depth is due to the size of Iceland which causes a narrow aperture of seismic network when compared to the size of the plume underneath. This problem has recently been overcome by Montelli et al. (2004a,b) based on an evaluation of finite-frequency travel time tomography for seismic wave. This technique is based on a wave approach, considering that travel time of a finite-frequency wave is sensitive to anomalies in a hollow, banana-shaped region surrounding the unperturbed ray path (Montelli et al, 2004a,b). Finite-frequency tomography images can be 30-50% larger than in other tomography analysis. Figure 4b shows the results from Montelli et al (2006). From these figure it appears that the Icelandic hot-spot does not have a P-wave anomaly in the lower mantle suggesting a plume origin at the 660km discontinuity. Contrastingly, an S-wave anomaly is observed from the surface to a 650km depth followed by no anomaly around 1000km and another anomaly underneath continuing to the 2800km core-mantle boundary. Montelli et al (2006) suggest that a pulsating behaviour of the plume may explain the shape of the conduit beneath Iceland.
It is noteworthy that Iceland and the NALIP satisfy almost all the prediction of the plume model; the plumes originate from a hot boundary layer (the core-mantle boundary). The plumes consist of a large head followed by a small tail (most of the igneous material was extruded during the first 2-3 million yrs). The flatten plume heads is of 2,400 km in diameter (Greenland volcanic passive margin) and both heads and tails erupted high temperature picrites. The last prediction of the model is that flood volcanism should be preceded by 500 to 1000 m of uplift and has yet to be demonstrated in the NALIP.
Seismic imaging of the mantle beneath Iceland has yet to provide an undeniable proof of a mantle plume. However geological and other geophysical data also converge toward that hypothesis. In the following section we explore the interaction of such mantle plume with the mid-Atlantic ridge.
According to models by Gripp and Gordon (2002) the absolute plate motion (motion compared to fix hotspots) is highly asymmetric. The Eurasian Plate is moving 1.4cm/yr towards the ENE while the North American Plate (NAP) is moving 2.7cm/yr toward the NW. This asymmetry causes the spreading center to move westward with time relative to the hotspot and has been ongoing since the opening of the north Atlantic. This hotspot track can be traced from Baffin Island and the west coast of Greenland 65Ma ago to the east coast of Greenland 40 Ma ago to the Westfjord area (Iceland) 20 Ma ago and finally to its present location under the Vatnajökull glacier (Figure 5a,c). From this observation it appears that the westward drifting of the NAP is slowing over time possibly explaining the higher topography of Iceland compared to the Greenland-Iceland and the Faeroe-Iceland ridges. Mittelstaedt et al (2008) modelled the mechanism of rift jump associated with a mantle plume. From their model of a fix hotspot beside a ridge axis, the ridge axis fully accommodates spreading at first while hotspot magmatism begins to thin and warm the off-axis lithosphere (Figure 5b). As the hotspot heating continues the off-axis lithosphere weakens, which initiates rifting and perturbs the mantle flow pattern, such that small amounts of upwelling begin to occur near the hotspot zone. Eventually, larger divergence rates develop and induce asthenospheric upwelling this is the beginning of the ridge jump and during that period coeval rifting occurs in both spreading center. Since the divergence rates at the two rifts are different, the lithosphere between them experiences small lateral velocity gradients, acting as a “microplate”. This microplate phenomenon is observable in Iceland where the area between the WVZ and the EVZ acts as a microplate bounded by large strike-slip fault systems. Finally, as the divergence at the original ridge axis ceases, the off-axis location establishes stable seafloor spreading.
Albers and Christensen (2001) study the interaction of mantle plumes with mid-ocean ridges using 3-D numerical convection models for different spreading rates and viscosity contrast. They found two end-member models which they baptised “Pipe flow” and “Pancake flow”. The “Pancake flow” case will develop if the lithospheric thickness variations are small compared to the plume layer thickness. Such flow occurs if lithospheric thickness variation is small because of a moderate or high seafloor spreading rate or if plume thickness is large because of moderate viscosities. The “Pipe flow” case will develop if the lithospheric thickness variations are appreciable compared to plume thickness. In this case the plume flow is sufficiently strong to inhibit flow away from the ridge axis and thus enhance flow along it. Such “pipe flow” appears to require both low spreading rates and very low plume viscosities (Albers and Christensen, 2001) and is possibly the closest analogue for Iceland.
Isotopic ratios are somewhatconstant along the “normal” segments of the mid-Atlantic ridge but vary greatly in spatial association with hotspots and fracture zones. Figure 6 shows the profile of several isotope couples along the mid-Atlantic ridge. From this figure it is clear that the La/Sm, 87Sr/86Sr and 3He/4He ratios
increase in Iceland rocks compared to the rest of the ridge. In term of the petrography the Cpx/Pl ratio also increases toward Iceland. The explanation for such isotopic variations is the mixing of distinct magmatic sources, namely Mid-Oceanic Ridge Basalts (MORB) and Ocean Island Basalts (OIB) (Ito et al, 2003). The high 3He/4He ratio is a strong indicator of a lower mantle component. 3He is a light element that was produced during the big bang and is not produced anymore; therefore the original 3He/4He ratio is in constant decrease. In Iceland, lava with gas trapped in vesicles gives 3He/4He ratio around 16 to 22, more than twice the ratio in MORB. This observation suggests a deep non-depleted mantle source to at least some of the lavas.
Now that we have a good understanding of the large scale tectonic processes that come into play in Iceland we can comprehend the complex geology that resulted from them.
The unique landscape of Iceland is the result of the interplay of volcanism and Glaciation. The volcanic activity is nowadays confined to the neovolcanic zone which represents the onshore extension of the mid-Atlantic ridge. This zone has been subdivided into rift zones (where extensive crustal spreading occurs) and flank zone (were little to no crustal spreading occur). Flank zones are characterized by alkali olivine basalt and transitional alkali basalt, thick crust and have a lower geothermal gradient. There are three volcanic flank zones; the Snaefellsnes Volcanic zone, the south Iceland Volcanic Flank Zone (SIFZ), and the Öræfi Volcanic Belt (Figure 7). Rift zones are characterised by Tholeiitic basalt, thin crust and high geothermal gradient. There are three volcanic rift zones; the West Volcanic Zone (WVZ), the East Volcanic Zone (EVZ) and the North Volcanic Zone (NVZ). The volcanic zones are connected by large transform systems known as fracture zones or volcanic belts when volcanically active. These zones are the Tjörnes Fracture Zone (TFZ), the Reykjanes Volcanic Belt (RVB) the South Iceland Seismic Zone (SISZ) and the Mid-Iceland Belt (MIB). Along the volcanic zones eruptions are contained within discrete portions termed central volcanoes many of which are associated with geothermal fields, silicic magmatism and calderas. Central volcanoes are associated with fissure swarms that display crater rows along normal fault systems. Central volcanoes together with these fissure swarms are termed volcanic systems (Saemundsson, 1978) (Figure7). Although no compression forces were applied in Iceland the volcanic succession displays large folds. These folds result from the weight of the volcanic system which depresses the crust bellow making the nearby succession dipping toward the volcanic center. This process has been operating since the Tertiary allowing paleo volcanic centers to be located.
As discussed above there are three magmatic series in
Iceland which spatial distribution is quite confined. Alkaline olivine rocks
make up the flank zones and are typically associated with stratovolcanoes.
Tholeiitic rocks constitute the rift zones and are mainly erupted by shield
volcanoes and their subglacial Tuyas equivalents. Transitional alkali basalts
are found where a flank zone and a rift zone converge toward each others.
Tholeiitic magmas are characteristic of MORB whereas Alkalic magmas are
typically associated with OIB or continental rift. Transitional alkali basalt
probably reflects a mixing of OIB and MORB signatures. Alkaline rocks represent
primary mantle melt; they are enriched in REE compared to MORB (this reflects a
lower amount of partial melting, concentrating the volatiles and explaining how
some can find their way to the surface quickly) and often contain mantle
xenolith. They are believed, in OIB, to represent material from the undepletted
lower mantle and typically occur at the beginning and at the end of hotspot
activity. Accordingly, the current activity along the off-rift Snæfellsnes
Peninsula represents dying hotspot volcanism unconformably overlying tholeiitic
lavas of Tertiary and Pleistocene age (clearly observable by the weathering
contrast between units), whereas the SIFZ has incipient (e.g. Surtsey) volcanism
related to the hotspot-driven southward propagation of the EVZ. From these
observations and from previous descriptions of rift jump mechanisms a
simplistic model can be proposed. At t=0 Alkaline magma are erupted from
hotspot-alone activity, at t=1 the rift joins the hotspot and transitional
alkali magma is erupted. At t=2 the rift activity gets concentrated, the degree
of partial melting increases (melting the OPX phase in the mantle) and only
tholeiitic magma is produced. At t=3 the
rift “jumps” to its next location and dying alkaline volcanism follows for a
The rock formations in Iceland are divided into four stratigraphic groups based on climatic and paleomagmatic conditions at the time of formation. This stratigraphy is divided into Tertiary, Plio-Pleistocene, Upper Pleistocene and Postglacial. Figure 8 describe the volcanic structures visited during this field trip in chronological order.
The rock formations older than 3.3 Ma constitute the Tertiary formation. These rocks cover about half of the island and occur away from the spreading zone with their age increasing toward the edge of the island. These rocks are found and have been observed in eastern, western and northern Iceland with the oldest rocks (15 Ma) found in the eastern and western extremities. Tertiary rocks formed before widespread glaciation and therefore seldom contain glacial deposits and subglacial volcanic rocks. Most of the Tertiary formations consist of monotonous (ca 10m thick) basaltic lava pile mostly subaerial separated by minor epiclastic interbeds and red-brown paleosoil which fossil plant remnants allow for climatic reconstruction (Figure XXX) (Thordarson and Hoskuldsson, 2002). The average lava deposition rate is of one lava flow every 10000 yrs. Extensive carving of the Tertiary during the last Ice age providing a cross section of volcanic systems that are analogue to the volcanic system seen along the neovolcanic zone at the surface today. These paleo-volcanic systems are characterized by numerous dyke swarms (equivalent to the fissure swarm at the surface today) and associated with andesitic to rhyolitic lavas and tephras.
Plio-Pleistocene rocks were not extensively looked at during this field trip, however they were shortly observed during the drive to Hekla as we drove through the Hreppar anticlinal and were seen at a distance from our Rekjavik bases camp as we looked NE to the Esja Mountain. These rocks share the same characteristic as the Tertiary succession on top of which they rest unconformably but also display subglacial successions such as móberg ridges and Table Mountains forming while the first large ice-sheet started fluctuating.
Upper Pleistocene rocks together with postglacial rocks make up the neovolcanic zone of Iceland and were observed at several locations such as the Landmanalaugar - Skogar to Vik area and around the Hengill central volcano. These rocks are characterized by increased subglacial volcanic edifices compared to interglacial successions. This indicates more extensive glaciations during that period (five glacial cycles) forming the characteristic móberg ridges and Tuyas Mountains such as the examined Hjorleifshofdi. This period also marks the formation of some sandur plains such as the Mýrdalssandur which built up by successive Jökulhlaup events and ice melting during interglacial periods.
The Holocene postglacial lavas have been observed at
numerous locations including the Vestmannaeyjar, rootless cones from the Eldgja
flow, Hekla and the Mývatn - Krafla area. These lavas are all subaerial and
fill topographic lows between the mountains of mostly Upper Pleistocene age. Soil
formation started covering the landscape as the ice retreated. Large glacial
river fed by the meltwater and episodic Jökulhlaup filled sandur plain. The
first settlers sailed to Iceland in 860 and recorded historical eruptions some
of which (e.g. Elgja 934-40) were examined during our field trip.
Postglacial and Upper Pleistocene lavas have been the primary focus of this field trip and will be extensively discussed in the next chapter on volcanism.
Iceland is a volcanologist paradise. The diversity of volcanic edifices, flows and textures makes this island a prime location to study volcanoes. In this chapter, after briefly introducing volcanic processes, we will present the different morphology of volcanoes seen in Iceland. Afterward we will present the different eruption products and finally spend some time on Ice-magma interaction.
Most of the different types of eruptive behaviour depend strongly on the composition and thus the temperature, viscosity and gas content of the magma as well as their replenishment rate and vent geometry.
The viscosity of magmas (its resistance to flow) depends firstly on pressure, temperature, water content and the magma composition. Shaw (1972) described the viscosity by the formula: Ln (V) = s.104/T –s . cT + cV (V=viscosity (in poises), s is a function of the chemical composition, T=temperature (in Kalvin), cT =1.5 and cV = -6.4). In this equation, water is taken into account by the factor s and decreases the viscosity quickly. The presence of phenocryst increases the viscosity because each solid particle in suspension is surrounded by a halo of liquid interfering with each other (“the phenocryst effect”). This effect is particularly important if the magma contains few large phenocrysts (McBirney and Murace, 1984). The density of a magma affects the floatability of phenocryst and hence the viscosity. Bubble in magma will first increase the viscosity in similar fashion than phenocryst. However as the bubble content increases, bubble blend and the viscosity decreases greatly. CO2 can combine with Na to produce Na2CO3 complexes polymerizing the melt and increasing viscosity (Schmincke, 2004) however magmas with high CO2 content such as Carbonatites are characterized by very low viscosity.
Numerous factors influence the triggering of a volcanic eruption. Tectonic controls such as large plate movement operate at a relatively long time scale. Magma mixing as new magma rise to the magmatic chamber can trigger eruption by inducing chemical and thermodynamic disequilibrium (Bardintzeff, 2006). This process is believed to be the driving force of recent eruption at Hekla, were resorbtion of the plagioclases provide evidence of such chemical change in the magma chamber. Volatiles (mainly water) from Juvenal or meteoric sources can overcome the solubility in the magma nucleating vesicles and increase its volume and hydraulic pressure. As the magma rises it becomes sursaturated in gasses which exolves crossing the glass transition and producing tufficites. Fractional crystallization processes also have an effect on magma eruptibility by increasing the volatiles content in the fluid phase each time a new mineral phase crystallizes. The presence of a resistant plug isolating the magma chamber from the atmosphere can result in the accumulation of enormous pressure which once liberated will draw the collapse of the whole edifice forming a caldera such as the Torfajökull and Krafla caldera.
Volcanic systems in Iceland form along fissure swarm en echelon networks of tensional cracks, normal faults and volcanic fissures. The most common volcanic type is fissure eruption. This type of volcanism occurs along narrow linear scars on the ground along which the magma is erupted resulting in a fire curtain. Such fire fountains are the result of a basaltic magma which degas easily. As the volatiles escape, they carry bombs and spatter toward the sky in a continuous fashion accumulating material along the fissure. The 1783 Laki eruption is a majestic example of such fissure eruption. A 25km long scratch and 115 eruptive vents were active for a period of 7 month producing 12.5km3 of lava. Associated lava flows covered an area of 565km2 (second largest historical flow on earth). The fire fountain at Laki reached an estimated height of 10 to 13 km reaching the base of the Stratosphere. Fire fountain of comparable sizes are rare but have been observed in the 1998 Etna eruption (1500m high for more than an hour) and the Izu volcano in the Oshima island in 1986 producing a 1600m high fountain. The volcanic landscape left by fissure eruptions is an alignment of cone named crater row centered along a large fissure (Figure 10). These are formed by larger fire fountains in the continuous fire curtain. Crater rows are characteristic of the Icelandic volcanic landscape and are seldom found anywhere else on earth. Such crater row have been observe at Heimaey where the 1973 Eldfell eruption produced a 5km long fire curtain made of closely spaced fire fountains before being reduce to a unique vent at the Eldfell scoria cone.
Point-source effusive volcanism is also quite common, forming lava shields which are produced by long-lived eruptions of olivine tholeiite or picritic pahoehoe flows that are fed by sustained lava lakes residing in the summit vent (Thordarson and Larsen, 2007). Lava shields volcanoes, such as the Petra volcano near the Hengill central volcano, form symmetric edifices with a gently sloping (3-10°) flank centered on a summit crater. Such volcanoes are typical of Hawaiian type volcanism; they are produced by surface flow and internal lava tubes (Figure 11).
Strombolian type activity is quite common in Iceland producing large quantities of basaltic scoria. In this type of system, lava is easily fragmented by the escaping gasses and little ash material is produced. The Eldfell volcano is an excellent example of Strombolian type activity since it concentrated the activity from the original fissure eruptions. Strombolian activity is characterized by crater-focused eruption columns of about a hundred meter high raining bombs and lava block that have sufficient time to cool in the air and not weld to each other as they fall. The result of such eruption is a scoria cone with a slope of ca 30° and accompanied by an A’a scoria basaltic flow (Figure 12). Strombolian eruptions can produce a large variety of bombs which spatial distribution is a factor of their distance from the vent. Figure 13 shows example of different types of bombs and detail on their genesis.
It is to be kept in mind that this classification is based on end-members and that in nature several eruption types can occur during the same eruption event or during the life of the volcano (e.g. Hekla is a composite stratovolcano and a fissure volcano). Surtseyan volcanic activities owe its name to the new born 1963 Surtsey Island in Vestmannaeyjar. It is the hydrovolcanic equivalent of Strombolian eruptions on land (Francis and Oppemheimer, 2004). Although Surtsey wasn’t visited during this trip; this type of activity is worth a diagram since it illustrates the early stage of all the Westman Islands (Figure 15). Surtseyan activity develops under shallow water (10s of meters) and is characterized by an unlimited access to water (hydrovolcanic not phreatomagmatic).
Along our description of the types of volcanic activity we portrayed several types of volcanic vents. However our dissertation overlooked many smaller or intricate volcanic edifices which we depict in the next section.
The spatter cone of Eldborg (Fortress lava) formed by accumulation of hot basaltic (low viscosity) magma ejected into the air for too short of a time to cool. As the still malleable lava falls back, it accumulates around the vent building circular vertical walls termed “spatter rampart”. The spatter is hot enough to be easily remobilized and the formation of a spatter cone is often accompanied by a pahoehoe flow.
Rootless cones form when lava flows over wet land or shallow water such as a lake a river or the shore. As the lavas enter in contact with the water it quenches and form glass that will insulate the rest of the flow from the water. Fracturing of this glassy bottom will result in a small phreatomagmatic explosion with the water under hydraulic pressure expelled up forming a mix cone of cinder and spatter. The term rootless refer to the absence of a true root system as the cones are fed by horizontally flowing lava tubes (Figure 16). Theses features are easily recognisable by their random pattern and were observed at several locations including the 934-40 Eldgjá flow rootless cones group and the Mývatn lake rootless cones. The size difference between these two examples is striking and reflects the amount of available water. The 2000 yrs old, Lake Mývatn cones were erupted while the magma was flowing over a lake, therefore lots of water was available to feed the explosion creating large cones of dominantly cinder material. The Eldgjá flow rootless cones were more in the magmatic side of a phreato-magmatic explosion as they flowed over wet land. Their water influx perished rapidly and only small cones of dominantly spatter material were created.
When magma is injected at shallow level and enters in contact with the groundwater a gas film develops that separates the magma from the water. When the film collapses, it fractures the rock; the water gets heated again and creates a vapor film once more that collapses and so on until suddenly a huge explosion is triggered. These explosions are called phreatic eruptions and get their explosive power from the Meteoric water, destroying the country rock. Such explosion results in maars which crater is surrounded by a tuff ring. This type of explosion occurred at the 1875 Víti double pumice crater (Krafla caldera) and at Hverfjall where it created a perfectly shaped tuff ring.
The 6500 yrs old Keriod crater is a collapse crater which started as a scoria cone. The crater was probably formed by a small magma chamber beneath the crater which emptied toward the end of the eruption resulting in a collapse of the cone centre. Spatter cones are made up of highly vesicular bombs and spatters which outer part is oxidized. These reflect localized volatile-rich basaltic systems where large gas bubbles burst throwing incandescent blobs of spatter into the air (Cas and Wright, 1987).
We have described the different volcanic landforms related to several types of volcanism. In the next section the focus is switched to the eruption products.
When large quantities of basaltic lavas accumulates in topographic depressions it forms levée which confine the flow suppressing any easy way to escape and forming lava lakes (or pseudo-lava lake as they are away from their roots) as seen in Dimmubrogir. These lava lakes are in constant motion isolated from the atmosphere by a carapace of solidified scoria under which the lava flows within a complex multilevel system of tubes. Dimmubrogir represent a magnificent site to explore the interior of a lava lake and all the features associated with lava tubes such as lava stalactite, skylight, and lava carpets (Figures 19 an 20). The surface of a lava lake shows numerous little fire fountain and is dynamic at the beginning of the eruption but freezes rapidly. The 1992-1997 formation of the lava lake in the Pu’u’O’o crater from the Kilauea volcano in Hawaii has been entirely recorded and gives an outstanding perception of what the surface of Dimmubrogir looked like 2000yrs ago (http://hvo.wr.usgs.gov/gallery/kilauea/). At Dimmubrogir the lava tubes where drained possibly following a breaking of the levée leaving collapsed structures. The opposite phenomenon occurs when liquid lava accumulates underneath a solidified crust which inflates and breaks open creating rounded monticules termed Tummulus. Several of the basaltic flow observed display very glassy texture. Large masses of glass cannot form by normal surface cooling (only the upper part of the flow should be glass) consequently, exsolution of the gas phase probably occurred changing the flow to glass (withdrawing the water and CO2 result in polymerization of the melt, hence turning it into glass if the gas phase escapes quickly). This is confirmed by the very high vesicular content of the observed rock. Another impressive structure observed at numerous locations of basaltic flows is columnar joint. These form by slow cooling of thick flows. Because the top and bottom of the flow cool before the central layer, these outer areas contract whereas the center doesn’t. This results in tensional stresses that creates regular joint sets as blocks pull away from one another to create polygons separated by joints. The joints propagate down from the top and up from the bottom as cooling progress toward the center forming columns (Winter, 2001). Columns form usually perpendicular to the surface of the flow (Figure 21) but their arrangement depends on the cooling surface. Excellent examples of such columnar joints can be found in the Gian valley where horizontal, rosette and chaotic columns are also found (cf: ice-magma interaction section).
Rhyolitic flows are very viscous (8-10 poise) and hence do not flow for great distances and are therefore good indicators of proximity to the vent (McPhie et al., 1993). They tend to form small domes or coulées. However, several coulées can build up into a larger edifice, which is the case of the 1300 meter high Torfajökull central volcano. This area is by far the largest area of exposed silicic extrusive in Iceland with more than 80% of surface outcrops being silicic in composition making it the largest silicic center anywhere on oceanic crust. (Gunnarsson et al. 1998). The volcano forms a large (450km2) caldera massif with an approximate volume of silicic extrusive of 225 km3 (Gunnarsson et al. 1998). Silicic rocks from the edge of the caldera in the Landmanalaugar area and from the 1477 Námshraun flow were observed. Figure 22 shows the features observed at Landmanalaugar. At this location a preglacial rhyolite of a marked green colour is entirely chloritized. This chloritization took place under greenshist facies metamorphic condition while the outcrop was under more than a km of ice and probably still heated by the nearby geothermal activity. The first assumption of metamorphic petrology is that it operates under isochemical conditions; therefore a high component of metasomatism is necessary for Mg and Fe to be moved into the rhyolite and chlorite to crystalise. Postglacial rhyolites have better preserved texture and show ubiquitous flow banding and some autobrecciation which occur as the highly viscous magma folds over on itself and blocks, breaks and get carried away by the flow.
Pyroclastic flows are a mixture of hot gas in expansion and rock material in suspension. They have been subdivided in three group; pyroclastic fall, flow and surge. Pyroclastic fall and surge deposits from the Hverfjall (“crater mountain”) tuff ring have been observed. Pyroclastic fall deposits are much more extensive than pyroclastic flow and surge deposits. They are not affected by topography and cover the landscape as a uniform blanket. However they are greatly affected by the predominant wind directions which dictate their distribution. Fall deposits are typically well sorted both vertically with finer particles toward the top and laterally with finer particles away from the vent (Winter, 2001). Also, they are rarely welded as they have time to cool in the air and are therefore not deposited while hot. Near the Hverfjall tuff ring there is further indications that these fall deposits were not deposited hot. In this region, pyroclastic fall was wet (mud rain). In hydrovolcanic explosion abundant steam in the pyroclastic clouds condense rapidly cooling the cloud (below 100°C) and sticky ash particles aggregate to form accretionary lapilli. The cohesiveness of the material results in plastering of even steep slopes with a veneer of fine ash (Francis and Oppenheimer, 2004). Near the Hverfjall tuff ring there is a succession of wet and dry deposits which reflect changing conditions such as access to water from the vent (i.e. recharging from groundwater) or weather conditions. Other indicators of wet deposition include desiccation cracks at the surface of some beds and the plastering of tree layer upon layer molding the tree without burning the bark. Typical dry surge deposits are hot and represent low density turbulent cloud of ash. Base surge are associated with hydromagmatic eruption and hug the ground as they propagate from the vent in all directions (similar to nuclear base surge). As they flow; they pick up gases which expend accelerating the flow and forming ash hurricanes. These flows are extremely destructive but yet leave very small traces in the geological record, recovering topography with a highly variable thickness. They often appear as fine grained thin little beds that habitually contain cross bedding (Figure 23). Other types of surge flows are associated with pyroclastic flows from which they can detach (decouple) and can even go over the topography. That is what happened in Mont Pele (Martinique) where in 1902, the city of Saint Pierre, 6 km away from the volcano and protected by deep valleys, got destroyed in a few seconds by such a flow also called a nuée ardente (La Croix, 1904).
Unconsolidated Pyroclastic fall deposits (tephras) together with reworked pyroclastic deposit (epiclastic) are observable at the Pumice quarry at the feet of the Hekla central volcano. At this location white pumice from the 1104 eruption overlays pink pumice from an earlier eruption and is partially reworked by later epiclastic deposits. Thin 2cm black tephras from the 2000 eruption cap the whole sequence. Hekla’s eruptions are associated with a large amount of meltwater from the snow at the summit which explains the abundance of epiclastic deposits. Pyroclastic rocks contain three types of fragments. Juveniles fragments are of similar composition as the matrix and are part of the same original magma, they are on the whole pumice fragments. Cognant fragments represent a portion of the magmatic vent, either part of the volcano that got picked up or of the magma chamber. Accidental fragments represent anything else that got picked-up along the way, observed accidental fragments include metamorphic rocks, baked up pieces of soil and wood. Juvenile and cognant fragments dominate at this location. Thermal oxidation and plastic deformation of fragments are found at the pumice quarry indicating hot deposition (Figure 24). Columnar jointing and welding are also proof of hot deposition and can be found at the welded rhyolitic tuff of Halanköjafell (no fiamme observed).
Volcanic gases are released in large quantities by some eruptions and are usually freed at the first stage of the eruption. The effect of volcanic eruption on the global climate has long been recognised (e.g . Forsyth, 1988) and their direct effect on health is a major concern. The Laki eruption produced noxious fumes emitted that stunted grass growth and killed more than half of the country’s livestock through fluorine poisoning. These consequences ultimately resulted in the disastrous "Haze Famine" that killed 20 per cent of the population (Thordarson, 2008). Ingested fluoride by man or animals initially acts on the intestinal mucosa. It can form hydrofluoric acid in the stomach which leads to irritation or corrosive effects (Nochimson, 2008). The Laki eruption injected 100 million tonnes of sulphur in the atmosphere into plumes that were dispersed eastwards over the Eurasian continent and north into the Arctic. Sulfur is released in the atmosphere as SO2 where it combines by photochemical oxidation with H+ to create sulfuric acid (H2SO4). About 175 million tonnes of those aerosols were removed from the atmosphere by subsiding air masses within high-pressure systems forming the infamous dry fog that hung over the Northern Hemisphere for more than five months. The acidity of the dry fog was such that it caused considerable damage to vegetation and crops all over Europe and stunted tree-growth in Scandinavia and Alaska (Thordarson and Hoskuldsson, 2002).
Common gases released during volcanic eruption include SO2, CO2, H2O, Cl, F and NOx. Sulfur is the gaseous species which profound effect on atmospheric composition is the most widely accepted (e.g. Rampino and Self, 1984). H2SO4 and dust absorb incoming solar radiation and increase the opacity of the atmosphere. The emitted dust and aerosol droplets will serve as nuclei in the atmosphere promoting the formation of clouds and increasing the albedo of the Earth. This will have for effect a global cooling event termed volcanic winter. Sulphuric acid together with chlorine and fluorine (which turn to HCl and HF respectively) will result in acid rain. The eighteenth century temperature records from the northern hemisphere suggest that the annual cooling that followed the Laki eruption was about 1.3°C and lasted for two to three years. Satellite-borne sensors are able to measure SO2, ozone, aerosols and atmospheric temperature providing strong direct evidences for a connection between individual explosive eruptions, the generation of volcanic aerosols, and climatic cooling. Volcanic gases from the 2000 Hekla eruption were sampled in-situe by plane. A volcanic sulfate aerosol cloud has a potential residence time in the stratosphere of up to several years but most of the fine volcanic ash lifted into the stratosphere by eruption columns settles out of the atmosphere in less than 3 months. This short-lived atmospheric residence presumably limits its influence in time. However, Bay et al. (2004) suggest that volcanic eruption could induce mineral fertilization of part of the ocean resulting in phytoplankton bloom which effect would not only be to consume CO2 but also to increase ocean albedo by as much as 10% contributing to a global cooling and possibly leading to a glaciation.
Glacial periods starting in the plio-Pleistocene and through the Upper Pleistocene were marked by the presence of large ice-sheet covering Iceland and resulted in the formation of noticeable volcanic structures. In the next section we explore the mechanisms of ice-magma interactions.
As stated above, basaltic point source and fissure eruptions are the most common in Iceland. The subglacial equivalent of these eruptions will form topographic features which morphology is mainly a function of the original glacier thickness (=hydrostatic pressure at the vent) and the internal volatile pressure in the magma. The heat transfer to the ice during subglacial volcanism is so efficient that the magma enters a subaqueous environment in the form of a water-filled ice cavity or ice-dammed lake (Allen, 1980). For thin glacier (<150m) the hydrostatic pressure on the vent is weak and the result will be a hydrovolcanic eruption somewhat similar to Surtseyan type producing a tuff ring and then being filled by till deposits. For thick glaciers (>150m) the hydrostatic pressure is greater so any explosive activity is prevented in a first time and sea-floor type effusive eruption with formation of pillow lavas and then pillow breccias being emplaced. As the ice melts and the edifice builds up, more explosive-type hydrovolcanic activity takes the lead with formation of vast quantities of tuff and hyaloclastites via a phase of pillow breccia. Most Icelandic subglacial volcanic mountains comprise cores of pillow lava, overlain by pillow breccia and hyaloclastite tuff, reflecting decreasing hydrostatic pressure as the mountain grows higher during the eruption (Trönnes, 2002). If the activity stops at this point, staying under the ice, the resulting landform will be a Móberg ridge (typically in subglacial fissure eruptions) as seen along the road in the eastern part of the Reykjanes peninsula. If the eruption continues however the edifice can build up high enough to reach the top of the glacier and continue its activity in a typical effusive subaerial fashion. When the glacier retreats it leaves a mountain with a flat top called Table Mountain or Tuya (Figure 25). The height of Table Mountains in Iceland is used to infer the thickness of the last glacier. It now appears that this view on Tuyas formation is a bulky simplification. Most Table Mountains are not formed by a single long-lived eruption but also show period of little activities with lacustrine sediments from the large sub-glacial lake including turbidites and debris flow (Schmincke, 2004). The chaotic facies of Tuyas include a mixture of scoria, breccia and pillow lava in a palagonite matrix. Palagonite is a dusty indurated yellow to brown texture of volcanic glass hydrated by thermal shock with cool water. Hjorleifshofdi (the cave outcrop) is an isolated Table Mountain that sits on the Myrdalssandur which shows a beautiful example of the facies succession in a Tuya. At this site however the ever-building sandur covered most of the initial pillow-lava stage. This mountain was isolated from the mainland at the time of the first settlers but was incorporated by the growth of the sandur (outwash plains) by successive Jökulhlaup and is now 2.5km away from the sea. The typical wave erosion structures on the large caves within this mountain testify of its past as island. Móberg ridges and Table Mountains have also been identified on Mars where they probably formed by a similar process (Chapman, 1994).
Jökulhlaups caused by geothermal or volcanic activity are the most common volcanic hazard in Iceland (Gudmundsson et al., 2008). Jökulhlaup (glacier burst) are sudden gush of water, ice and rock debris from the meltwater of a glacier above an erupting volcano. Icelandic glaciers are wet-based so water is ever-present at the base of the glacier. During or shortly after an eruption the newly created subglacial lake opens a draining path and all the other meltwater joins it, creating for a few hours to days the single highest quantity of moving water on the planet. Along Hgw1 on the Myrdalssandur the ruins of a bridge destroyed by the 1996 Jökulhlaup (formation of the Gialp volcano) can be seen. This Jökulhlaup occurred five days after the beginning of the eruption and released 3 km3 of water at a rate of 45 000m3/s! The water travelled under the ice for a distance of 50km. Recent retreat of the Vatnajökull glacier allows seeing the resulting Esker which formed in a day. This finding challenges classical understanding of Esker creation (seen as long-lived sub-glacial rivers). The outwash plains (sandur) grow by these 6-10 years interval Jökulhlaups, transporting volume around 5 km3 of rock.
Large ice-sheets do not only modify the type of volcanic activity and their deposits but can also shape the amount of volcanism. During the unloading and crustal rebound period following the ice-sheet retreat, magma eruption rate increased by a factor of 30 to 50 relative to present eruption rate (Maclennan et al., 2002). This increase has been documented all along the neovolcanic zone (Slater et al., 1997; Sigvaldason et al., 1991; Vilmundardottir and Larsen, 1986) and is due to an increase in mantle melting rate by decompression melting following the removal of the Icelandic ice-sheet.
We described in some length the volcanic processes associated with sudden eruptions. However volcanic-related activity also occurs in a more continuous fashion over longer periods of time. In the next chapter we describe the geothermal environment surrounding volcanic centers.
Geothermal environments regroup all phenomenons due to water with a temperature significantly higher than the regional norm. In Iceland geothermal areas are present near most central volcanoes such as Torfajökull, Krafla and Hengill. Geothermal areas are commonly subdivided into high (>150°C) and low (<150°C) temperature. High-temperature areas are only found along the neovolcanic belts or their periphery while low-temperature areas are found all over the country, except in the East and South-east where they are less abundant. All geothermal manifestations are the result of meteoric water heated bellow the vadose zone and going back to the surface without cooling significantly. A hydrothermal system can be viewed in a simplistic way as a large convection cell. The meteoric water infiltrates into the rock and gets heated by a magmatic body at depth causing the water to rise. Some of the water is expelled through hot springs and related systems while some of the water moves horizontally and get cooled by the newly infiltrating meteoric water. These hydrothermal fluids will cause hydrothermal alteration of rocks by removing and adding components. The effect of such process is beautifully exposed by large mountain faces exhibiting colourful tint for instance in the Krafla caldera area.
In this chapter we will explore the use of geothermal water as an energy source, describe the surface features of geothermal field and discuss briefly the life thriving in that environment.
Icelanders have used geothermal heat for centuries but this use was primarily limited to bathing and laundering. Stefán B. Jónsson is considered to be the first Icelander to heat his house with water from a hot spring while methods of heating were already used in France for centuries. Hot water distribution started in 1930 and by 1945, 2850 houses were connected to the power grid (Orkustofnin, 2009). Since that time the development of geothermal energy has been impelled to match the demand of power-intensive industry and at present faces new large scale projects such as the possible extensive production of hydrogen as a synthetic fuel.
Geothermal energy exploitation is surprisingly simple and can be divided into three phases; the collection and processing of steam from boreholes, the heating of cold water and the production of electricity (Figure 26). Boreholes collect steam mixed with water which is sent through collection pipes to the separation station where the water is separated from the steam by gravity. Excess steam and unutilised water go into a steam exhaust outside the separation station. The steam is then dried and its temperature lowered to 190°C before it enters the turbine with a pressure of 8 to 12kbar. It is that pressure difference that makes the turbine turn producing electricity. Each turbine produces 30 MWatt of electricity. The steam is then cooled in the condenser by heating freshwater to 60°C, condenses to water and is injected back to the ground. The geothermal water heats the freshwater in a tube heat exchanger without mixing. The freshwater heated by the steam and the one heated by the groundwater mix and is sent to the deaerator where boiling under low pressure releases the dissolved oxygen and other gases from the water. To eliminate the last traces of oxygen and lower the pH, a very small quantity of steam containing hydrogen sulphur (H2S) is added to the freshwater giving it its characteristic smell. The freshwater at 82°C is then channelized to the cities through insulated pipelines.
Geothermal energy is a hard resource to manage because the quantity is unknown. It is a renewable energy though, clean compared to most others and in large availability in Iceland. Exploration for geothermal energy involves calculating the heat discharge of an area which can be done by air reconnaissance just at the end of winter looking at melted areas. It also involves steam sampling of CO2 H2S He and H ratios from hot springs to estimate the reservoir temperature. Evidently an understanding of the local geology and stratigraphy is essential to develop model of groundwater flow and place boreholes.
The epitome of any geothermal area is Geysers. They are the
result of superheated water trying to turn into steam and pushing from beneath
and being counterbalanced by the weight of the overlying water column. This
makes the geyser oscillate up and down. Eventually the upward pressure
overcomes the column weight; some of the water is pushed to the side giving
less resistance to the rising overheated water which turns to steam and degas
in a quick blow. Such a balance is a
rare phenomenon (compared to hundreds of hot-pool and fumaroles). Geysers need
a heat source 3-5 km from the surface and a porous rock substrate (ex: connected
joints or glacial sediments) to assure a high recharge rate of the reservoir. As
the meteoric water infiltrate in the ground it is quickly heated by the high
geothermal gradient and can dissolve minerals easily, especially silica. At the
surface the silica precipitates as opal then crystallize as cristobalite and
quartz forming layers of Geyserites (hydrous amorphous silica). Before 1935
Geysir was depositing 10cm of geyserite every twenty years. Eventually the silica accumulation will plug
up the system. The expected lifetime of a geyser is about 1000 yrs. Geysir itself was visited and stayed calm as
its neighbour Strokkur (the clock) propelled every few minutes.
Solfataric fields are home to a variety of hyperthermophiles bacteria and algae and their submarine equivalents are thought to represent the closest organism to the first form of life. Liquid water is fundamental for hyperthermophiles and boiling hot springs are too hot for any form of life. At lower temperatures (or higher pressure) hyperthermophiles can prosper, they only rely on simple inorganic compounds and grow between 80 and 113°C. The distribution of hyperthermophiles species in the solfataric field is mainly a function of the heat gradient and complex microcosm of hyperthermophiles are present even at the highest temperature end. Hyperthermophiles are chemolithoautotroph and use mixture of oxidized and reduced minerals and gases as an energy source to assimilate carbon from CO2 (they are independent of sunlight). They are anaerobic since oxygen has a low solubility at high temperature and its penetration in the soil is prevented by CO2 (Stetter, 2005). In hydrological terms volcanoes can be viewed as “hot sponges” containing lots of aquifer likes cracks and holes and might represent an undiscovered environment of hyperthermophiles biotas (Stetter, 2005). Hyperthermophiles therefore only need liquid water, simple chemical compounds, heat and anaerobic conditions to thrive and propagate. The early earth atmosphere was reduced and volcanism was much widespread (Valley, 2006) uniting all the requirements for hyperthermophiles life not only near black smokers but also on land in solfataric fields. It is therefore possible that hyperthermophiles life was dominant and wide-spread in the early Archean. Evidence from the moon of a “Late heavy bombardement” period of frequent meteorite impact at about 3.85 to 3.9 (e.g Koeberl, 2006) raise an interesting hypothesis. Ejectas produced during impact can be expelled out of the planetary system and hyperthermophiles can survive for long time at temperature up to -140°C in a dormant state. It is therefore tangible that hyperthermophiles have been able to colonize other planets such as Mars which was volcanically active at the time. The presence of Martian meteorites on Earth testifies of the material exchange between the two planets and therefore life’s beginning could have originated in any of the two terrestrial planets.
Iceland is to geology what France is to wine and cheese. It is the place where geology textbooks turn to life displaying world-class examples for every discipline in the earth sciences. In this report we have focused on the volcanology of the Island and have been able to describe all major types of volcanic activity within the mere 100km2 of the Island. Volcanic landforms, structures and textures form under specific conditions which reflect the magma viscosity and gas content. Observing such features in active volcanic environments displaying clear relationships is indispensable to unravel ancient volcanic successions. Most eruptions in Iceland involve basaltic magmas and are therefore not very explosive. This suggests at first glance that the danger for the population is limited. However Jökulhlaups, lahars and volcanic gases can represent real threats to the population. In recent history however volcanoes have represented less of a natural hazard than earthquakes and storms in Iceland. This statistic can be misleading to decision-makers and should be balanced by the destructing effects of the 1783 Laki eruption. This eruption killed 20% of the population at the time and its effect on air transit nowadays would be catastrophic. The effect of volcanic eruption on high altitude aircraft was a myth before a brand new Boeing 747 lost 4 engines going through the 1990 Redoubt volcano in Alaska. In order to monitor volcanic activity the Science institute at Rekyavik operates a network of between 30 and 40 short-period seismic stations distributed throughout Iceland. Recently much attention has been attributed to the Hekla volcano where a new monitoring method based on water level from boreholes is being developed.
Allen, C.C. 1980. Icelandic subglacial volcanism: thermal and physical studies, J. Geol. 88, 108-117.
Allen, R. M., Nolet, G., Morgan, J.W., Vogfjord, K., Bergsson, B.H., Erlendsson, P., Foulger, G.R., Jakobsdottir, S., Julian, B.R., Pritchard, M., Ragnarsson, S., and Stefansson, R. 2002 Imaging the mantle beneath Iceland using integrated seismological techniques, J. Geophys. Res., 10.1029/2001JB000595
Bardintzeff, J.M. 2006. Volcanologie. Dunod
Bay, R.C., Bramall, N., Price, P.B. 2004. Bipolar correlation of volcanism with millennial climate change. Proceedings of the National Academy of Science of the United States of America 101, 17, 6341-6345
Buck, W.R., 1991. Modes of continental lithospheric extension. Journal of Geophysical Research 96, 20161– 20178.
Campbell, I.H. 2005. Large Igneous Provinces and the Mantle Plume Hypothesis. Elements, 1 p265-269
Cas, R.A.F., and Wright, J.V., 1987. Volcanic Successions: Modern and Ancient: London: Allen and Unwin.
Chapman, M.G. 1994. Evidence, age and thickness of a frozen paleolake in Utopia planitia, Mars. Icarus, 109, 393-406
Corti, G., Bonini, M., Conticelli, S., Innocenti, F., Manetti, P., Sokoutis, D., 2003. Analogue modelling of continental extension: a review focused on the relations between the patterns of deformation and the presence of magma. Earth-Science Reviews 63 169–247.
Francis, P. and Oppenheimer, C. 2004.Volcanoes. Oxford university press.
Gorski, P., 2009, The rediscovery of the liverworts Anastrophyllum donnianum and A. saxicola in Central Europe (Slovakia, Tatra Mountains): Cryptogamie Bryologie, v. 30, p. 409-414.
Gripp, A.E., and Gordon, R.G., 2002, Young tracks of hotspots and current plate velocities: Geophysical Journal International, v. 150, p. 321-361.
Gunnarsson, B., Marsh, B.D., and Taylor, H.P., 1998, Generation of Icelandic rhyolites: silicic lavas from the Torfajokull central volcano: Journal of Volcanology and Geothermal Research, v. 83, p. 1-45.
Gudmundsson, M.T., G. Larsen, Á. Höskuldsson and Á.G. Gylfason. 2008. Volcanic hazard in Iceland. Jökull, 5: 251-268
Hirose, K. 2007. Post-perovskite phase transition and the nature of the D’’ layer. In Yuen, D.A., Maruyama, S., Karato, S., Windley, B.F. 2007. Superplumes: Beyond Plate Tectonics, Springer
Hopper JR, Dahl-Jensen T, Holbrook WS, Larson HC, Lizarralde D, Korenaga J, Kent GM, Kelemen P.B., 2003 Structure of the SE Greenland margin from seismic reflection and refraction data: Implications for nascent spreading centre subsidence and asymmetric crustal accretion during North Atlantic opening. Journal of Geophysical Research 108(B5): 2269, doi:10.1029/2002JB001996
Ito, G., Lin, J., and Graham, D., 2003, Observational and theoretical studies of the dynamics of mantle plume-mid-ocean ridge interaction: Reviews of Geophysics, v. 41.
Jóhannesson, H. and K. Sæmundsson, 1998. Geological map of Iceland, 1: 500.000. Bedrock geology. Icelandic Institute of Natural History and Iceland Geodetic Survey, Reykjavík.
Koeberl, C. 2006. Impact Processes on the Early Earth. Element, 2, 211-216.
Maclennan, J., Jull, M., McKenzie, D., Slater, L., Grönvold, K. 2002. The link between volcanism and deglaciation in Iceland, Geochem. Geophys. Geosyst., 3(11), 1062
McBirney, A.R. and Murace, T.1984. Reological properties of magmas, Annu. Rev. Earth. Planet. Sc., 12, 337-357.
McPhie J, Doyle M, Allen R., 1993. Volcanic Textures: A Guide to the Interpretation of Textures in Volcanic Rocks. Hobart, Tasmania, CODES, Univ. of Tasmania, 196 p.
Mittelstaedt, E., and Ito, G., 2005, Plume-ridge interaction, lithospheric stresses, and the origin of near-ridge volcanic lineaments: Geochemistry Geophysics Geosystems, v. 6.
Mittelstaedt, E., Ito, G., and Behn, M.D., 2008, Mid-ocean ridge jumps associated with hotspot magmatism: Earth and Planetary Science Letters, v. 266, p. 256-270.
Montelli, R., G. Nolet, F. A. Dahlen, G. Masters, E. R. Engdahl, and Hung, S.H. 2004a. Finite-frequency tomography reveals a variety of plumes in the mantle, Science, 303, 338–343
Montelli, R., G. Nolet, G. Masters, F. A. Dahlen, and Hung, S.H. 2004b. Global P and PP traveltime tomography: Rays versus waves, Geophys. J. Int., 158, 637–654
Montelli, R., G. Nolet, F. A. Dahlen, G. Masters, E.R. 2006. A catalogue of deep mantle plumes: New results from finite-frequency tomography. Geochemistry, Geophysic, Geosystems. 7, Q11007, doi:10.1029/2006GC001248, 2006
Nicholson, H., Condomines, M., Fitton, J.G., Fallick, A.E., Grönvold, K., Rogers, G. 1991 Geochemical and isotopic evidence for crustal assimilation beneath Krafla, Iceland. J. Petrol. 32, 1005-1020.
Nochimson G. 2008. Toxicity, Fluoride. eMedicine, 12-28.
Orkustofnin, J. 2009. Energy development in Iceland. Lecture at the Iceland Geosurvey.
Oskarsson, N., Sigvaldason, G.E., Steinthorsson, S. 1982. A dynamic model for rift zone petrogenesis and the regional petrology of Iceland, J. Petrol. 23, 28-74.
Rampino, M.R. and Self, S., 1984. The atmospheric impct of El Chichon. Scientific American, 250, 48-57.
Saemundsson, K. 1978. Fissure swarms and central volcanoesof the neovolcanic zones of Iceland. Geological Journal Special Issue, 10, 415-432.
Saunders, A.D., Fitton, J.G., Kerr, A.C., Norry, M.J., and Kent, R.W., 1997. The North Atlantic igneous province. In Mahoney, J. J., and Coffin, M. F. (Eds.), Large Igneous Provinces. Geophys. Monogr., Am. Geophys. Union, 45-94.
Self, S., L. Keszthelyi and T. Thordarson, 1998. The importance of pahoehoe. Ann. Rev. Earth Planet. Sci. 26, 81–110
Shaw, H.R. 1972. Viscosities of magmatic silicate liquids: an empirical method of prediction. Amer. J. Sc., 272, 870-893.
Sigvaldason, G.E., Annertz, K., Nilsson, M. 1992. Effect of glacier loading/deloading on volcanism: Postglacial volcanic production rate of the Dyngjufjoll area, central Iceland, Bull.Volcanol. 54, 385–392.
Slater, L., Jull, M., McKenzie, D., Gronvold, K. 1998. Deglaciation effects on mantle melting under Iceland: results from the northern volcanic zone. Earth and Planetary Science Letters 164 (1998) 151–164
Schmincke, H.U. 2004. Volcanism. Springer
Stetter, K.O. 2005. Volcanoes, hydrothermal venting and the origin of life. In Marti, J. and Ernst, G. (Eds),Volcanoes and the environment. Cambridge University Press.
Thordarson, T.. 2008. Historical Flood Lava Eruptions The 1783-84 Laki and 934-40 Eldgjá events. IAVCEI General Assembly field guide.
Thordarson, T. and Hoskuldsson, A. 2002. Classical geology in Europe 3, Iceland. Terra Publishing
Thordarson, T. and G. Larsen, 2007. Volcanism in Iceland in Historical Time: Volcano types, eruption styles and eruptive history. J. Geodyn., 43, 1, 118–152.
Trönnes, R. G. 2002. Introduction on the geology and geodynamics of Iceland: Nordic Volcanological Institute, Reykjavík. <http//www.norvol.hi.is
Valley, J.W. 2006. Early Earth. Elements, 2, 201-204
Vilmundardottir, E., Larsen, G. 1986. Productivity pattern of the Veidivotn fissure swarm, Southern Iceland, in postglacial times. Preliminary results: paper presented at 17e Nordiska Geologmo¨tet, Helsinki.
White, R and McKenzie, D. 1989. Magmatism at rift zones: The generation of volcanic continental margins and flood basalts, J. Geophys. Res., 94, 7685-7729
Winter, J.D., 2001. An introduction to igneous and metamorphic petrology. Prentice Hall.