Continental rifting of the Late Paleozoic supercontinent of Pangaea, progressive break-up and initial spreading of the Atlantic Ocean was preceded by the intraplate emplacement of large igneous provinces in the Mesozoic-Paleogene. Fragmented by seafloor spreading, eroded remnants of the dissociated continental flood basalt events are distributed on formerly conjugate margins of rifted and drifted landmasses across the Atlantic domain.
The break-up of the boreal region of Pangaea between Greenland and Europe led to opening of the North Atlantic beginning in the Paleocene following the formation of the North Atlantic Igneous Province. It too was fragmented by seafloor spreading between the North American and Eurasian tectonic plates on opposing sides of the diverging Mid-Atlantic Ridge.
The entire northern Pangaea event and more - LIP emplacement, extension, rifting, break-up and seafloor spreading - is widely attributed to the buoyant arrival of the hot, deep-seated Iceland mantle plume. Its interaction with the mid-ocean ridge produced persistent voluminous and effusive hotspot magmatism that culminated with the formation of Iceland.
The rate of magma production along the Mid-Atlantic Ridge has remained unusually high. Iceland, at the center of the northern North Atlantic Ocean, is the only part of the North Atlantic Igneous Province that remains magmatically active, albeit at a reduced flux rate.
An alternative genetic origin to the plume hypothesis, one that is gaining ground, relies solely on shallow kinematic processes. In this scenario, plate tectonics, a nearly universally accepted process, drives lithospheric extension that permits the sub-lithospheric mantle to passively ascend and generate surface magmatism at a "meltspot."
Regardless of the formative mechanism, the result is the elevated Iceland Basalt Plateau, the world's largest volcanic island. Young geologically, it's thought to have formed within the last 24 million years in spite of the fact that the oldest exposed rocks on the island are only 14 to 16 million years old.
It’s one of the most active and productive terrestrial regions on Earth. There are some 55,000 km of diverging plate boundaries on our planet, and along most of their length, they are submarine features. Iceland is one of only two places where a mid-ocean spreading center rises above the level of the sea.
Iceland is a geological paradise with an on-land continuation of the Mid-Atlantic Ridge, volcanism that is subaerial, subglacial and submarine, at least 120 historical eruptions in the last 1,100 years that represent 60% of all recorded volcanic activity, ubiquitous Tertiary flood basalts and an eruption frequency about every five years.
Although undeniable, Iceland borders on the inexplicable. Few places are as pristine, remote, otherworldly, serene and stunningly beautiful. It’s a sought after destination for its breathtaking landscapes and the unparalleled opportunity it affords to observe magmatism, active tectonic processes on land and their interaction with climate.
If you seek to comprehend the fascinating history, culture and heritage of Iceland, then understanding its geology is essential. If you seek to better understand the natural forces that shape our planet or simply contemplate the enigma of mantle structure and geodynamics, it's the place to be.
ABOUT THIS POST
The photographs in this post were taken by my daughter Julia Share (designated as such) and me on two separate visits to Iceland with the exception of a few contributions from a friend and colleague. Our excursions began in the capital district of Reykjavik (arrow) and included the Golden Circle region, the Reykjanes Peninsula, the South and Southeast Coastal Lowlands, and the Snæfellsnes Peninsula. Our plan is to return and negotiate the 1,332 km-long Ring Road that encircles the island and allows access to Iceland's rugged, isolated and largely uninhabited highlands and more remote geological features.
Relevant definitions in this post are italicized and important names are emphasized in boldface. Global coordinates of select locations are provided and can be pasted into a mapping program such as Google Earth. Click on images and maps for a larger view. Voluminous additional references beyond those provided are available upon request.
The Golden Circle region is on everyone's must-see list. Northeast of Reykjavik, it includes Iceland's famous and celebrated rift valley, national park and UNESCO Site of Þingvellir, geothermal fields with hot springs, mud pots and geysers such as Geysir and Strokkur, the massive and famous waterfall Gulfoss and enigmatic Kerið crater.
Southwest of the capital extends the elongated Reykjanes Peninsula. It's a landscape of volcanic fissures, lava fields, fault-bound lakes and geothermal activity. It includes Keflavik International Airport, the man-made Blue Lagoon spa, the Svartsengi power station and four volcanic systems arranged en echelon. The peninsula is a UNESCO Global Geopark with 55 listed sites, lying on strike with the Mid-Atlantic Ridge as it rises from the sea.
The South and Southeast Coasts are regions of coastal lowlands, sea cliffs, sea crags and caves, countless waterfalls and expansive glacial outwash plains that extend to the sea. They provide access to a number of active volcanoes, south highland glaciers and associated geomorphological features, the largest of which is glacier Vatnajökull and National Park, one of three in Iceland and Europe's second largest.
North of Reykjavik, after following zig-zagging fjords along the coastal plain, is 80 km-long, 10-30 km-wide Snæfellsnes Peninsula in West Iceland. "Iceland in miniature" has some of everything to offer geologically. Dominated by dormant stratovolcano Snæfellsjökull and its eponymous glacier and National Park, the diverse region includes Iceland's most photographed Kirkjufell mountain and scenic fishing villages that cling to the coast.
Our late May weather included everything from impenetrable fog and pelting, icy rain to cloudless, brilliantly blue skies, which is a brief commentary on Iceland's variable climate, travel preparedness and the never-ending photographic challenges one encounters.
A few tips: Bring rain and wind-proof layers for blink-of-an-eye weather changes. Summers are beautiful but not hot. Bring a bathing suit for geothermal pools found everywhere. Look beyond the high-tech but very communal, touristy Blue Lagoon. Icelandic króna (ISK) is unnecessary except for small purchases and some rural areas, but bring those credit cards. Venture out of Reykjavik! Hire a guide with a SuperJeep or rent a car or 4X4 for backcountry F-roads (gravel roads with river crossings). Off-road driving is forbidden. Everyone speaks English. Campgrounds are plentiful. Bring a sleep mask for the midnight sun. The food is good but expensive. Mobile connections are spotty on the road.
WHERE IS ICELAND?
It's in the middle of the northern North Atlantic roughly between Greenland, Scandinavia and the British Isles. After Great Britain, it's Europe's second largest island, slightly larger than Ireland. With an area of about 103,000 sq km (40,000 sq mi) only 30% is above sea level. If you take the shallow, circumferential shelf into account, the dimensions swell from 300 x 500 km to 450 x 750. Owing to the large numbers of elongate fjords that were carved into its peripheral landscape, Iceland's coastline is an astounding 6,090 km-long.
Contrary to popular perception (or just geographic confusion), Iceland is part of Europe, even though it's divided between two tectonic plates. In reality, so is coastal California, part of North America but vertically sliced by the North American and Pacific plates.
HOW ÍSLAND GOT ITS NAME
It's a common myth that the "ice-land" was named to deter enemies in pursuit and encourage seafarers to continue on perhaps to a seemingly more hospitable, greener Greenland. It was not only discovered and settled after Iceland in 985 but was penned by Norwegian Viking Erik the Red, remembered in medieval and Icelandic saga sources, to actually encourage visitation and settlement (unless that too is a myth).
Previous known names for Iceland were Snæland or "Land of Snow" by Naddoddur Ástvaldsson, the purported Norse discoverer of Iceland, and Garðarshólmi or "Isle of Garðar" by Garðar Svavarsson, a Swedish Viking who briefly resided there. So, who actually named the country and why?
Icelandic sagas (ancient manuscripts written in 'Old Norse') tell of western Norwegian seafarer and explorer Flóki Vilgerðarson, who settled in Iceland in 865 during the Viking Age (800 to 1066 AD). Nicknamed Hrafna-Flóki, which means "Raven Floki" for the birds he released to find land, he cursed the country after a particularly harsh winter after failing to gather enough fodder (winter food) for his starving cattle. He was prevented from departing by an ice-choked fjord and therein gave the country the name that finally stuck.
First to briefly occupy Iceland were Gaelic monks seeking isolation and religious freedom. According to two ancient books - one by an Icelandic priest in the early 12th century and by Landnámabók, the Norse's Book of Settlements in the 9th and 10th centuries, settlement in earnest began in the second half of the 9th century when Norse settlers (Germanic inhabitants of Scandinavia) began migrating across the North Atlantic. The accursed name "Land of Ice" did little to deter Ingólfr Arnarson, Iceland's first known permanent settler and founder of Reykjavik in 874.
HERITAGE, TRADITION AND LANGUAGE
The Icelandic language of Íslenska is derived from 'Old Norse', brought over by Norwegian Vikings and influenced by Scandinavian and northern European neighbors. Without dialects and having many runic alphabet letters from Germanic languages before the adoption of the Latin alphabet over a millennium ago, it's barely changed in some 800 years. What's more, spoken and written purity are protected by the Icelandic Language Institute that approves every new word such as tölva for computer, which means "number-prophetess".
"Icelandic Tongue" is celebrated on Language Day every November 16th. It's an example of nationalism and independence that allows sagas from the earliest days to be read in the language they were written to help preserve ancient culture and heritage. Íslenska is the most basic element of the national identity, but connectivity and globalization are taking their toll with words like "OK" that have crept into the lexicon.
Like other Germanic languages, Íslenska combines difficult to pronounce words into longer ones. Geologic and geographic prefixes and suffixes are perfect examples: -jökull meaning glacier; -foss, waterfall; -eldfjall, volcano; -jökulsá, large river; -höfn, harbor; -ey, island; -fells, isolated mountain; -breidur, broad volcanic shield; -vatn, lake; -fjöll, mountain; -hraun, lava; -fjallgarður, mountain range; -öskjur, caldera; laugar-, hot spring; reyk-, steamy; -dalur, river valley; -vellir, field; -vegur, road; -falljöklar, icefall; -stapi, table mountain or tufa; -sandur, glacial outwash plain and -ur, for place names.
As a result, Eyjafjallajökull is the "English Island-Mountain-Glacier"; Snæfellsjökull, the "Snow-Mountain-Glacier"; Dimmifjallagardur, the "Dark-Mountain Range"; Reykjavik, the "Smoky Bay", Laugarvatn, the "Hot Spring Lake", and Surtsey, the island volcano that is the genitive case of the mythological Norse giant of fire.
LAND OF FIRE, ICE, RAIN, SLEET AND SNOW
The Arctic Circle passes through the tiny Icelandic island of Grímsey, and yet, the mainland enjoys a moderate maritime climate. With small seasonal variations in temperature due to the north and west-flowing Irminger Current, a branch of the North Atlantic Current and northeast extension of the Gulf Stream, it delivers relatively warm, high-salinity Atlantic waters and plentiful moist air to the southern and western coasts.
Persistent precipitation in the southern and central Highlands, where snow accumulation exceeds ablation (loss from melting, sublimation and evaporation) fuels glacier formation in a tundral climate. They're remnants of the vast Iceland Ice Sheet that blanketed the island to the shelf break and northern Eurasia in the Pleistocene. Today, volcanoes display two shapes depending on the age of generation: plateau-shaped formed during glacial periods and shield-shaped if formed in the post-glacial Holocene.
During the Last Glacial Maximum, the last time ice sheets were at the greatest extent in Late Weichselian-time of the late Pleistocene and early Holocene. It's thickness exceeded 2,000 m and covered mountain summits. It extended as far as the shelf break doubling the size of the island, evidenced by ice-contact landforms such as 100 km-long and up to 50 m-high terminal moraines on the sea floor over 100 km around West, South and East Iceland.
Before glaciation blanketed the landscape of Iceland, Miocene through Pliocene lavas were the dominant landform. In the Pleistocene, major glaciation influenced landforms in volcanic zones across the landscape with hyaloclastite ridges and mountains (essentially melted cavities within ice) and steep-sided table mountains (the same with extruded lava on the surface). Both formed subglacially (or submarine), whereas the former emplaced from an elongate fissure and the latter from a central vent.
Spurned by climate deterioration and controlled by rising global seas, due in part to melting of ice sheets such as the Laurentide in North America and the Eurasian, rapid deglaciation began between 18.6 and 15 k before the present.
Iceland's glaciers are long-standing reservoirs of ice that turn to meltwater dependent on climate and locale. Released water enter the subsurface to feed aquifers and source rivers, the largest and most energetic of which are harnessed for hydropower. It accounts for more than ~75% of domestic energy production, while geothermal energy is used for heating and electricity.
Glacial runoff, which is frequently enhanced by jökulhlaups (catastrophic outbursts from subglacial eruptions, geothermal areas and sudden release of ice dammed lakes) may carve gorges into the successions of lava that built the volcanic island and blanket the landscape and spill over countless rapids and waterfalls leaving deposits across broad sandur outwash plains.
Holocene warming caused rapid deglaciation, controlled by rising global sea level due to melting of ice sheets such as the Laurentide in North America, which began between 18.6 and 15 k before the present. Ever since, glaciers have been retreating and re-advancing (due to climate deterioration such as the Dryas) inside the coastline. Distinct shorelines formed, represented by sea cliffs along the central South Coast.
By 8.7 k, deglaciation had obliterated the Eurasian and Iceland Ice Sheet, and segregated Iceland's ice sheet it into some 269 named glaciers at higher elevations in the Highlands. Their location mirrors the maze of volcanic zones and belts that have been forming for over 16 million years. The result is a landscape that's about 11% glaciated, while overall, snow and sleet account for ~7% and ~35% of annual precipitation. Indeed, Ísland - EES-lahnd in Icelandic - is most deserving of its appellation.
BAROMETERS OF CLIMATE FLUCTUATION
Succumbing to climate change, Okjökull in West Iceland, which has been severely diminished by ablation, lacks the sheer mass to move under its own weight. In 1890, it was 16 sq km in size and is the first to no longer qualify as a glacier, simply known as “Ok” (rhymes with 'talk'). Future climate scenarios indicate even larger glaciers will follow suit in 150-200 years (some say as much as 500), perhaps with the exception of those on the highest peaks.
Melting of Icelandic ice's 3,600 cu km of water would raise global seas by 1 cm. Regardless of the time frame, it's the current trend seen in Iceland, as backwasting (peripherally), downwasting (on the surface), the terminus (toe-end of the glacier) and firm line (that separates the zone of accumulation and ablation) retreat upvalley. It is predicted that the Arctic Ocean could be ice-free mid-century as well. Prescient-thinking, lowly Reykjavik on the coastal southwest has even begun to reassess its harbor infrastructure.
ISOSTATIC GLACIAL REBOUND
Counterintuitively, sea level at Iceland has dropped as its landlocked glaciers have melted. It's due to isostatic rebound of the landscape following glacial unloading of the lithosphere, a normal process of deglaciation. Parts of Iceland, especially on the southwest Reykjanes Peninsula, gradually relieved of the massive weight of ice some 2,000 meters thick and extending out to sea far enough to double the size of Iceland, are rising faster than any place on Earth - as much as 1.4 inches per year.
As the crust rebounded, it carried the landscape with it, exposing the volcanic shelf as the new coastline, while former sea cliffs became stranded inland from their original location and markedly uplifted, appearing as if formed by erosion or fault scarp uplift. Evidence is along Iceland's South Coast that occurred in the warmer Holocene beginning ~13,000 years ago when sea level rose some 100-150 m above the present level. The escarpments are known for its spectacular waterfalls that source from highland glaciers.
Interestingly, crustal rebound appears to have triggered earthquakes and pulses of volcanic activity as subsurface pressures and stress adjustments act on magma chambers and melt generation, especially if shallow. It's a reminder of the interrelatedness of Earth's processes.
Almost every type is found in Iceland - ice sheets (continental-size masses not necessarily associated with mountains), ice caps (miniature sheets), ice fields (even smaller), ice streams (slowly-moving and ribbon-like), outlet glaciers (tongues that drain a sheet or cap), surge (fast-moving and short-lived), alpine (many over active volcanoes and geothermal areas), valley (originates from mountains), piedmont (spilling onto a plain), cirque (bowl-like) and tidewater (recurring advance and retreat with calving).
By far, at 8,300 sq km, Vatnajökull in the southeast is not only Europe's largest ice cap - covering ~13% of Iceland's surface - but the largest outside the polar region. It has over 30 outlet glaciers that flow outward centrifugally, each constrained by valleys and troughs they've created.
Hidden beneath is a diverse landscape of U-shaped valleys, canyons, rivers and lakes, and seven volcanoes. It's also the location of Europe's largest National Park. Some of Iceland's largest rivers originate in Vatnajökull such as Jökulsá á Fjöllum to the north that, on its way to the Greenland Sea, has carved Jökulsárgljúfur, the country's largest canyon with famous waterfalls Dettifoss and Selfoss.
WHAT IS ICELAND GEOLOGICALLY?
It's the world's largest, solitary volcanic island, not to be confused with volcanic island chains along interplate subduction zones (eg. Antilles arc in the Caribbean and Aleutians in the North Pacific) or intraplate chains (eg. Hawaiian-Emperor Seamounts). It's a basalt plateau - an elevated geological feature of relatively low relief but considerable overall elevation (over 4,000 m above the seafloor and emerging to an elevation of 1,775 m), crustal thickness (3 or 4 times thicker at 10 to 14 km) and area of 350,000 sq km.
It was built by vigorous and effusive magmatism that repeatedly flooded the landscape with largely basaltic lava flows that accumulated over time. A subject of considerable debate is how magmatic productivity allowed a mid-oceanic ridge on which Iceland resides to emerge from the seafloor and construct an elevated basalt plateau above the level of the sea.
Iceland's sharply-delineated volcanic passive margin differs from the broad shelf of its Atlantic neighbors such as the North American East Coast that constitutes a passive continental margin, marked by thermal subsidence, massive sedimentation and seaward dipping lavas (flood basalts). Both types of margins are related to continental break-up, but in Iceland, it is commonly thought to occur over a hotter mantle with a high rate of lithospheric extension and is related to the development of a large igneous province.
UBIQUITOUS "ICELAND STONE"
Iceland's rocks are less varied from those of other regions and consist mainly (>90%) of volcanic basalt. It's the building block of Iceland generated mainly from rift zones. Sedimentary rock comprises only 5-10% and is greatest after the end of major glaciation. With the exception of rock formed at the margin of basaltic intrusions such as dikes, metamorphic rocks - formed under high pressure and temperature - do not occur.
Basalt typically has an aphanitic (fine-grained) texture due to rapid-cooling at the Earth's surface, enough to allow crystalline growth yet slow enough to form dramatic, vertical colonades that are mainly hexagonal in cross-section if the lava flow is sufficiently massive. Lava surface morphology is either pahoehoe (shiny, smooth, glassy, ropy and thinner), a'a (rubbly, slower and thicker) or blocky (even thicker).
Basalt is a dark-colored, mafic igneous rock - a ferro-magnesian silicate rich in Mg, Fe, olivine, pyroxene and plagioclase. Its low-silica and low gas content (45-52% SiO2) in a partially molten state (984 to 1,260°C) results in low viscosity (fluidity). That allows basaltic lavas (extrusive magma) to be non-explosive and flow considerable distances (tens of km) at variable speeds (6 to 30 mph) from its source (a single vent or long fissure), all facilitated by a high rate of effusion (eruption rate) and steeper topography.
Gabbro is the intrusive (subsurface), chemically-equivocal, phaneritic (coarse-grained, slow-cooling) version found in shallow magma chambers and exposed dikes. Less common in Iceland are intermediate and explosive-erupting felsic igneous rocks (light colored, high SiO2, feldspar, quartz and muscovite), which is called a full spectrum or bimodal association when occurring together.
As a result of basalts physical and chemical properties, countless blanketing successions of basaltic lava are underfoot, cover the landscape and form hyaloclastite ridges and table mountains.
Table mountains or tuya are basalt flow-constructed. Ubiquitous across the landscape, many formed as a submarine intrusion but most over an elongate volcanic fissure (móberg ridge) under glacial ice during the Pleistocene, when Iceland was covered by an ice sheet. As geothermal heat melts a confining ice cover, it forms a subglacial lake with signature pillow lavas that convert to brecciated hyaloclastites (glassy, quench-fragmented rock in a fine-grained matrix). When the eruption conduit reaches the surface, pahoehoe lava typically blankets the flat-topped, steep-sided edifice.
LARGE IGNEOUS PROVINCES
The Iceland Plateau is the mid-ocean centerpiece of a Large Igneous Province (LIP), a large accumulation of igneous rock, referred to as the North Atlantic Igneous Province (NAIP) that spanned the North Atlantic from Greenland to the northern British Isles (Morton and Parson, 1988). LIPs are intraplate magmatic events that formed relatively rapidly within oceanic and continental environments in a few million years, although some persisted for tens of millions extruding immense volumes (>100,000 cu km and often >1,000,000) of mainly mafic magma.
LIP emplacement is distinct and separate from magmatism associated with plate-boundary seafloor spreading and subduction events. They're an essential process in shaping our planet that precede continental rifting (extension, break-up and subsequent seafloor spreading). Beginning in the Proterozoic but likely in the Archean, Earth history has been punctuated by these crustal provinces (regions with common geomorphic, structural, temporal and genetic attributes) but are best preserved in the Mesozoic and Cenozoic. They include flood basalts, volcanic rifted margins, oceanic plateaus, giant dike swarms, etc.
LARGE IGNEOUS PROVINCES OF THE ATLANTIC REALM
Three discrete Mesozoic-Paleogene LIPs emplaced before break-up of the late Paleozoic supercontinent of Pangaea and the subsequent opening of the Atlantic Ocean. On the present-day landscape, the once-unified LIPs exist as fragmented remnants vastly reduced in size by erosion and distributed across the ocean by tectonic fragmentation on pre-rift conjugate margins of the continents of the Atlantic domain:
• the North Atlantic Igneous Province (NAIP) that opened the North Atlantic ~55 Ma between Greenland and Europe with Iceland at the center;
• the Central Atlantic Magmatic Province (CAMP) ~195 Ma between northeastern North America and northwest Africa;
• the Paraná-Etendeka Igneous Province (PEIP) between South America and southwest Africa that opened the South Atlantic ~120 Ma.
Similarly, NAIP volcanism and methane degassing of seafloor sediments into the atmosphere appears to be associated with short-term warming but an extinction event of uncertain magnitude, as is the PEIP.
GALLERY OF NAIP, CAMP AND PEIP REMNANTS
The following photos depict various remnants of the NAIP, CAMP and PEIP on conjugate continental margins that were separated by seafloor spreading and plate divergence. Again, Iceland forms the volcanically active centerpiece of the NAIP, while the other LIPs are magmatically extinct. What might account for this volcanological phenomenon?
GEOLOGIC SETTING OF THE NAIP AND ICELAND
Iceland, representing the active center of the NAIP, is situated at the juncture of two elongate seafloor structures of volcanic origin: the Mid-Atlantic (MAR) and Greenland-Iceland-Faeroe Ridges (GIFR). The interpretation of the relationships of these features with Iceland is the crux of a longstanding debate not only about the formation of the volcanic island but the structure and behavior of the Earth's mantle, and not just in the North Atlantic but globally!
THE GIFR AND MAR
The Greenland-Iceland-Faeroe Ridge is a roughly W-E trending submarine aseismic ridge or rise. With abnormally thick oceanic crust, it is a mountainous chain of seamount-volcanoes that extend from Greenland through Iceland (GIR segment) to the Faeroe Islands (IFR). It doesn't produce seafloor spreading or earthquakes except at its end beneath Iceland, where the theorized head of a mantle plume is centered.
Straddling the Mid-Atlantic Ridge, Iceland is the longest area of subaerially exposed mid-ocean ridge on the planet. The spreading center is a NS-trending, 10,000 km-long, volcanic mountain chain from Bouvet Island near South Africa to 330 km short of the North Pole and bisects the ocean floor between the drifted continents of the Atlantic realm.
It's offset (laterally-interrupted) by relatively short transform faults. A consequence of plate tectonics on a sphere, they are strain-relieving fracture zones that convey mainly strike-slip motion (horizontal, side-to-side) that neither create nor destroy lithosphere. The Jan Mayen Fracture Zone is on the north and Bight FZ and Charlie-Gibbs FZ are on the south of Iceland. The former's history of eruptive magmatism is thought by some members of the scientific community to be part of the mantle plume that formed Iceland or a fragmented micro-continent beneath Iceland.
The submarine MAR is a slow-spreading-type divergent plate boundary between the North American-Eurasian and South American-African plates that are separating at the rate of about 1.8 cm/yr. The mid-ocean spreading center is also a constructive plate boundary, where most of the Earth's crust is generated in diverging conveyor-belt fashion (red arrows). Over 70% of all spreading centers in the world's mid-ocean ridge systems are oriented obliquely to the direction of absolute plate motion.
To complete the concept, existing crust is removed via subduction at destructive margins, while plates that meet at a transform fault margin move parallel to each other called conservative margins.
|Schematic of Mid-Ocean Seafloor Spreading and Volcanism|
Modified from whoi.edu
On a global scale, the MAR is part of a 80,000 km-long, mid-ocean seafloor-system that includes a volcanically-active 65,000 mile-long mountain range, the longest in the world. It encircles the planet like the raised seems on a baseball with crests that rise on average 1,000 to 3,000 meters above the adjacent ocean floor. Mid-ocean ridges are found in every major ocean in Earth. Normally, they don't build up above sea level with the exception of Iceland athwart the MAR in the North Atlantic.
The MAR and GIFR ridges, fracture zones and NAIP - played a role in the evolution of Iceland and the North Atlantic, but not everyone agrees on how. One widely accepted interpretation is challenged by several alternative hypotheses - one of which that is gaining ground. Let's briefly explore that notion and then return to a discussion of Iceland.
A NEW GEOLOGICAL PARADIGM
During the Plate Tectonic revolution of the 1960's that accounts for volcanism at
interplate boundaries, geoscientists sought an explanation for exceptions such as
intraplate volcanism of the Hawaiian-Emperor seamount chain (within the Pacific plate, thousands of miles from the nearest plate boundary of any kind, read about it here) and the persistent and voluminous interplate volcanism of Iceland.
The solution was provided in 1963 by Canadian geophysicist-geologist J. Tuzo Wilson's Hotspot hypothesis. In it, the uindirectional, time-progressive volcanic island chain formed as the Pacific oceanic plate migrated over the relatively small, stationary melt anomaly of a hotspot (persistent hot region of the underlying Earth's mantle) where magma continuously breaks through the lithosphere. All that was needed was a mechanism of heating that came eight years later in 1971 when geophysicist Jason Morgan enumerated the Mantle Plume hypothesis.
THE MANTLE PLUME HYPOTHESIS
As originally defined, there were about 20 mushroon-shaped plumes of long-lived, fixed (stationary relative to one another) diapirs (narrow, cylindrical heat-flow conduits), each with a pendulous tail and bulbous head. Less dense, solid but not molten, they are theorized to originate at great depth (~2,892 km) from the core-mantle boundary (the Mohorovicic discontinuity or Moho, a thermal boundary) to which they anchor ("deeply-rooted").
The unusually ("anomalously") hot, buoyantly upwelling material ascends through the uppermost mantle to the surface and, flattening against the more rigid lithosphere, fuels a region of hotspot volcanism via adabiatic melting (partial melting of a portion of solid mantle with a different resulting composition) from decompression (rapid decreased pressure like a champagne cork popping open).
The result is the extrusion of basaltic magma onto the surface. The process commonly occurs at divergent tectonic plate boundaries, such as at mid-ocean ridges but also explains volcanism in association with continental lithospheric extension (e.g. Basin and Range and Yellowstone). Of course, Iceland is unique in that a mantle plume is thought to be rising up through the Mid-Atlantic Ridge. In fact, the concept of a plume rooted in the deep mantle beneath Iceland dates back to the definition of Morgan's hypothesis.
|Simplified Schematic of the Hotspot-Plume Concept|
1.) Rising mantle plume anchored at the core-mantle boundary; 2.) Decompression melting and 3.) Surface extrusion of basaltic volcanic rocks. From wmblogs.wm.edu
Simple and elegant, it made sense and found near universal acceptance. It integrated well with Plate Tectonic theory while providing its driving mechanism, a process for cooling our planet, recycling of the crust and explaining hotspot melt anomalies. Replacing Hall and Dana's Geosynclinal Theory of 1859 - Earth cooling and contraction that ultimately builds mountains - it has served the scientific community well for over 50 years.
Unified with tectonic theory, the concept of hotspot-plumes explains a world of possibilities regarding crustal thinning and extension, flood basalts, rifting, seafloor spreading, large igneous provinces and their distinctive isotopic signatures.
THE SAGA OF THE ICELAND PLUME - A PLUME-RIDGE INTERACTION
In the case of Hawaii, the magma source is from the top of a plume, fixed relative to the mantle below the migrating Pacific plate that produces a time-progressive track. In the North Atlantic, hotspot-plume advocates espouse Iceland's elevated and thickened oceanic crust and unusually high level of persistent volcanism is due to the interaction of the Mid-Atlantic Ridge spreading center and the Iceland plume that feeds the hotspot.
The plume's arrival beneath the migrating plate formed the GIFR hotspot track, the trail of volcanism indicative of where the plume had been (see dated track on map above). In essence, the Iceland hotspot was 'captured' in the MAR spreading system following NAIP break-up that formed as North America and Greenland drifted away from Europe following the rise of the plume.
A "PLUMELESS" ALTERNATIVE EXPLANATION
Plume skeptics or "platists" (Foulger, Natland, Anderson et al, 2001), using lines of argument that "plumists" use as confirmation, find fault with inconsistencies within the plume hypothesis. Contending that first-order observations fail to exist, only imaging, modeling and indirect chemical assays, proposed plumes don't meet the standard criteria to be called as such.
They say the extreme variability of hotspots and their tracks can't possibly fit into a "one-size-fits-all" model to explain them. Furthermore, they contend that hotspots aren't any hotter than normal and plume isotopes (such as helium) can be generated at shallow depths. As for Iceland's volcanism, it's remained entirely on the MAR since the opening of the Atlantic at ~54 Ma, not tracking along a GIFR hotspot time-track.
Platists believe Iceland is a "natural consequence of shallow processes related to plate tectonics" and that lithospheric spreading and thinning permit melted rock to escape to the surface. Thus, melting anomalies - hotspot volcanism - are the result of passive rather than active processes that occur where lithosphere is in extension along lines of structural weakness within the crust such as transform zones, old sutures and faults.
They further contend that their plumeless model accounts for every major observation at Iceland - geophysical, geological, petrological, geochemical and seismological. They assert that the concept is consistent with plate tectonics and doesn't have to invoke an endless barrage of "ad hoc plume variants" of every imaginable shape, size and number. For example, some plume models - such as cactus and spaghetti plumes - have acquired an extremely complex geometry and internal composition to refine and accommodate any given situation.
THE ICELAND "PLATE MODEL"
Platists believe Iceland began to form where the MAR - the line of opening of the Atlantic Ocean that formed at ~54 Ma when Greenland and Eurasia rifted apart - crossed a transverse outer branch of the relict Caledonian suture on the Atlantic seafloor at shallow depth (lower right panel). The suture is the site of a ~440 year-old subduction zone created when what are now Greenland and Scandinavia collided as the intervening Iapetus Ocean closed during a phase of the multi-stage formation of Pangaea.
Iapetus closure (part of the Wilson Cycle) occurred when three Paleozoic continental cratons (stable crustal blocks) collided - Laurentia (the Greenland region of ancestral North America), Baltica (ancestral Scandinavian micro-continent) and Avalonia (an elongate micro-terrane of volcanic islands that rifted from Gondwana as a prelude to the formation of Pangaea). For clarification, Rheic Ocean closure occurred when Gondwana
(South Hemispheric Paleozoic mega-continent) collided with Laurasia (Laurentia + Baltica and Avalonia) to form Pangaea.
Awaiting formation of the MAR, lithospheric faults and old sutures such as the Caledonian provide pathways for the passive ascent of decompression-melted oceanic crust from the upper mantle. The event built the GIFR crustal band from Greenland to Britain over the suture and explains the persistently large melt volume, tholeiitic geochemistry (from partial melting), since the suture remnants are entrapped and can't migrate laterally.
ONE MANTLE WITH MULTIPLE INTERPRETATIONS
The formation of the North Atlantic provides clues to the origin of hotspot-meltspot globally. Where lithosphere is in extension, Platists contend that passive volcanism formed the Hawaiian-Emperor chain due to an intraplate extensional stress field within the Pacific plate. In contrast, the Plume model postulates the Hawaiian plume - where Tuzo's theory was first proposed - is responsible for the formation of the volcanic island chain as the Pacific plate migrates over the hotspot.
To which mantle dynamic do you subscribe? Are both concepts mutually exclusive? Are we experiencing the beginning of a paradigm shift on a tectonic scale? Let's return to Iceland where the Mid-Atlantic Ridge is on land.
THE MID-ATLANTIC RIDGE ON LAND - A COMPLEX OF VOLCANIC AND SEISMIC ZONES
As the submarine MAR connects with Iceland, it forms the submarine Kolbeinsey Ridge (KR) in the Arctic Ocean in the north and rises on land as the Tjörnes Fracture Zone (TFZ), while the Reykjanes Ridge (RR) in the southwest rises as the Reykjanes Volcanic Belt (RVB). On land, their extensions, which are postulated to have been continuous during Iceland's formative stages, converge on the complex that has been reorganizing eastward since the volcanic island emerged from the sea and, as one might expect, is
interpreted antithetically by plumists and platists.
The EVZ is the most active volcanically and is an axial rift in the making (more on that later) that will eventually transition from the WVZ, where there's no longer volcanic activity (the last eruption was about 1100 AD, although the time frame within the last 10,000 is considered dormant). There is, however, an abundance of spreading and faulting such as at Þingvellir and geothermal activity in the region.
VOLCANIC SYSTEM - THE PRINCIPAL GEOLOGIC STRUCTURE IN ICELAND
Episodic crustal spreading allows swarms of vertical dikes to deliver magma to the surface and feed some 30 active volcanic systems within the volcanic zones. With a lifetime of about 0.5 to 1.5 myr, they are the principal geologic structure in Iceland.
When present, eruptive activity within each system - called fires or "eldur" in Icelandic - is focused within a central volcano, often caldera-capped, glacier-covered and with more differentiated magmatic products (not just basalts but intermediate and felsic rocks), and/or fissure swarms (5 to 20 km-wide and 50 to 200 km-long) that consist of elongate, subparallel, deep tensional crustal cracks and normal faults.
With activity closely linked to plate movements, volcanoes and fissures are surface expressions of shallow or deep-seated crustal magma chambers.
Every eruption type occurs in Iceland determined by magma composition, temperature and locale of emplacement: Surtseyan (violent and water-contact explosive in shallow seas, Strombolian (gas-driven mild blasts), Hawaiian (effusive, voluminous, fountaining basalts) to Plinian (massive, gas-driven and stratospheric).
Depending on viscosity (fluidity), gas content, contact with water (submarine and subglacial) and delivery vehicle (conduit or fissure) to the surface, Iceland displays all known volcaniforms. Examples include: shield volcanoes (low profile, lava-built, symmetrical, long-lived, conduit-fed domes), lava domes (viscous mounds), stratovolcanoes (cone-shaped, conduit-fed, layered composite-built) and caldera (explosive summit collapse or extravasation) to spatter (gassy vent-ejected mound), scoria-cinder (steep-sided, tephra-built, short-lived, fissure-fed cone) and tuff cones and maars (hydrovolcanic fragmented-rock within circular depressions).
SYNCLINES, ANTICLINES AND UNCONFORMITIES PROVIDE GENETIC CLUES
The Snæfellsnes Rift Zone in West Iceland was actively spreading until ~6 Ma when the plate boundary transitioned to the east. Partial proof resides on the landscape where Tertiary strata synclinally dip and strike (point at and tilt towards) in the direction of the older spreading center and then reverse direction and partially underlie strata (with unconformities) of the newly-forming West Volcanic Zone to the east.
It can be seen along the SW-NE trending Borgarfjörður anticline, where strata (some of Iceland's oldest) are along the synclinal axis (downward linear fold). The identical landform (and similar transition process) is between the active West and East Volcanic Zones demonstrated by the Hreppar anticline between the West and East Volcanic Zones that follow the same axial trend as the rifts.
JUMPING RIFTS AND PROPAGATING RIDGES
Loading by volcanism tilts the strata towards the magma-generating volcanic zones forming a shallow syncline centered on the spreading axis and a shallow anticline in the region between volcanic zones. The arrangement tells the story of a "rift jump", where plate boundaries shift from one volcanic zone to another.
It's unclear whether the western spreading center relocated eastward in single episodes or by gradual propagation of new rift branches from the MAR. What is apparent is that rift segments exist in various stages of activity from nascent to established to waning over time. Thus, the Öræfi Volcanic Belt in East Iceland Belt is volcanologically nascent, while dormant Snæfellsjökull volcano in West Iceland erupted recently enough to be considered active, although it is though to be subsiding.
If we assume the plume that feeds the Iceland hotspot is a stationary structure, then the position of the spreading axis must have changed over time by migrating west-northwest, concurrent with North American-Eurasian plate divergence. Iceland's crust is in a constant state of transformation, as spreading centers and volcanic zones relocate progressively eastward in order to remain coupled (in proximity) to the Iceland hotspot. In the process, microplates (crustal blocks) transfer from one major plate to another.
In contrast, platists look to the shallow mantle, plate tectonics and extensional lithospheric stress for an explanation of eastward rift reorganization. If Iceland's oldest rocks in the extreme northwest at ~16 Ma and in the extreme east at ~13Ma, were spreading apart at ~2 cm/yr, then they would be separated at ~290 km rather than 500 km as they now are. Their conclusion is that ~210 km-wide older crust (oceanic or continental in origin with the latter possibly a remnant of a Jan Mayen micro-continent) formed earlier and underlies Iceland, submerged beneath younger lavas (and accounts for Iceland thicker crust).
Regardless of the mechanism that facilitates the delivery of anomalously hot mantle material, as older formerly active rifts become extinct (recall the Snæfellsnes Zone), new ones initiate. This has occurred several times in Iceland at ~24, 17, 7 and 3 Ma. In fact, Iceland is currently experiencing a rift-transition from the West Volcanic Zone (where Þingvellir is located) to the East Volcanic Zone, a rift-in-the-making where 80% of the volcaic activity is occurring. And historically, the Reykjanes Ridge in the southwest and Kolbeinsey Ridge in the northwest are thought to have been continuous before detaching any evolving into the rift complex to the east.
ÞINGVELLIR - ASSEMBLY FIELDS AND WORLD-CLASS GEOLOGICAL SITE
About 45 km northeast of Reykjavik is Iceland's most important and most visited site. The hallowed ground is the spiritual heart and soul of the nation and main attraction on the Golden Circle tour. It's where Icelanders celebrate their sovereignty and independence, where tourists flock by the car and busload, and where geologists pay tribute to Iceland's most dramatic geological feature. Þingvellir is a "protected national shrine for all Icelanders" according to documents that designated it a national park in 1928 and a UNESCO World Heritage Site in 2004.
Many central events in Iceland's history occurred at Þingvellir - notably the establishment in 930 of Alþingi (or "All-Men's Assembly") that became the National Parliament in Reykjavik in 1845, conversion to Christianity from the Norse pagan god Æsir in 1000 that ultimately provided national peace and unification, and in 1944, gaining independence from Denmark that resulted in the declaration of the new Icelandic Republic.
Although early Icelandic settlers most assuredly were unaware of the region's geological significance, they undoubtedly recognized something gomorphologically unique about it. It may seem off the beaten path from Reykjavik, but over a millennium ago its location was accessible and centralized from populated areas, its openness accommodated large numbers and its ledges afforded vantage points for regional chieftains to address the populous.
The geology of the region made Þingvellir an ideal location to conduct important political and commercial business, recite and make the laws of the land, voice concerns, settle disputes and socialize.
THE LAY OF THE RIFT
Þingvellir is a 4.7 km-wide graben (German for "grave"), a continuously (~4 mm/yr) subsiding (downdropping) tectonic depression, Iceland's deepest (70 m at the west) and widest (10 to 25 km). Subsidence is caused by compaction of volcanogenic material and lithostatic loading of erupted material.
Highly asymmetrical (deepest immediately to the east of east-facing faults along the western margin), southwest-sloping, NE-SW striking and partially filled by the waters of lake Þingvallavatn to the south, the graben or rift valley formed as rifting (lithospheric extension) - the consequence of tug of war between two tectonic plates - episodically stretched and opened the landscape some 5-10 mm/yr across the central axis of the West Volcanic Zone.
Today, it accounts for 20-30% of extension across South Iceland, the rest taken up by the East Volcanic Zone, in transition as previously discussed. And, as we know, the large scale deformation and thinned lithosphere produces seismic and magmatic activity manifested in the region's earthquakes and volcanic systems.
As a whole Iceland is spreading at ~18.9 mm/yr at the plate boundary in the north (where there's the single rift of the NVZ) and ~20.2 in the southwest (more complicated with oblique spreading on the RVZ and two parallel WVZ AND EVZ but primarily across the latter). Again, it's assumed to be due to the presence of the Iceland mantle plume presently beneath Vatnajökull.
At the northern end of the Hengill Volcanic System, the rift valley developed and subsided between the diverging North American and Eurasian tectonic plates. Within boundary faults on the west (Almannagjá) and east (Hrafnagjá), the extensive lava flow-blanketed, stream-dissected, grass-vegetated graben floor is scored with fissure swarms (sub-parallel tension gashes), bound on the south by lake Þingvallavatn, Iceland's largest (83 sq km) and deepest (114 m) lake, surrounded by four volcanic systems. In the last 9,000 years, the total amount of downward displacement is over 40 meters with 70 meters of extension.
An old Icelandic proverb states, "Frjósöm er vatn sem liggur undir hrauni" or "Fertile is water that runs under lava." As such, 83 sq km and 144 m-deep, lake Þingvallavatn is Iceland's largest. Residing in the southern part of the Þingvellir graben, its basin formed by tectonic rifting, faulting, subsidence and glacial erosion and was modified by volcanic activity. Its waters are exceptionally clear, clean and icy cold (3-5°C). Although the nutrient concentration is low, it provides fertility for many plants and fish such as trout and char and exceptional visibility for divers and photographers (below).
The lake is sourced from groundwater springs, lava-filtered aquifers fed by precipitation and river Öxará (the Axe River, possibly called as such since weapons were ritually thrown in during the Alþingi) that originates from glaciers as far as Langjökull 50 km to the north.
TRANSITION OF THE WVZ TO THE EVZ
Þingvellir is located at the Hengill triple junction, a three plate intersection of two rift zones and a transform zone - the Reykjanes Ridge oblique rift, the South Icelandic Seismic Zone and the Western Volcanic Zone. Three plates meet there: the North American, Eurasian and the Southern Iceland or Hreppar microplate (crustal block named after the central volcano), between the WVZ and EVZ. Þingvellir records rifting along the WVZ, which was the primary zone of spreading between ~6 Ma and 2 Ma when the EVZ formed from propagation from the NZV.
Today, only 20-30% of spreading in southern Iceland is accommodated by rifting on the WVZ, while the remainder is on the EVZ. The implication is that plate spreading at Þingvellir is transitioning eastward, rift jumping through the volcanic/seismic zone complex that has developed. The phenomenon clearly exists, but interpretations of the process responsible for it are the subject of debate among geoscientists and are related to mantle and plate geo-dynamics.
Not only does rifting along the West Volcanic Zone demonstrate a decline of volcanism, but the ultra-slow spreading rate of the Þingvellir graben, its fast subsidence and extraordinary depth have led many to suggest that the WVZ is a failed or dying rift. It fuels the implication for some is that the rift is magma-starved and that all motion will eventually be transferred to the EVZ as volcanism is converting to stretching and normal faulting.
THE GENETIC BIG PICTURE (OR HOW OLD IS ICELAND?)
Without collecting field data, drawing geological maps or conducting tectonic analyses, a simple mathematical calculation will provide the answer to the question of Iceland's age. If measured along its west-east spreading-direction from shelf to shelf (not shore to shore), Iceland is about 500 km-wide. Since the spreading rate is about 2cm/yr, Iceland must have taken around 25 million years to form in the Early Miocene.
This is in spite of the fact that the oldest exposed crustal rocks are Middle Miocene in age. They appear as fine-grained Tertiay Basalts (as previously discussed) in the extreme northwest (~16 Ma) and east (~13 or 14 Ma). They are located furthest from the volcanic zones that formed them, while progressively younger rocks are found as one moves toward the central spreading axis. And (as also discussed), more than one age-progression set exists, since it appears that at one time there was more than one spreading axis that transitioned eastward. The age distribution of Iceland's formations tells us something about the genesis of Iceland, axial divergence and jumping rifts.
In the big picture, the formation of Iceland and the North Atlantic is the consequence of:
• the emplacement of the North Atlantic Igneous Province, which appears to have formed in two phases: a failed rift in the Labrador Sea in the "Middle" Paleocene (~62-58 Ma) between between Greenland and North America proper and in the latest Paleocene to earliest Eocene (~56-53 Ma);
• the extension and rifting apart of boreal Pangaea beginning around 200 Ma between Greenland and Eurasia beginning ~90 to 150 Ma;
• the opening of the North Atlantic in the Early Eocene (initiated ~54 Ma) and syntectonic development of the Mid-Atlantic spreading center.
Somewhere around 24 to 25 Ma, plumists believe the Iceland plume rooted in the deep mantle was the driving force in the process that formed Iceland when it contacted thinned lithosphere of the Mid-Atlantic Ridge. For them, the Greenland-Iceland Ridge on the Atlantic floor is the trace of the plume head.
In contrast, anti-plume platists, that deny the existence of a plume and presence of a hotspot track, adhere to a tectonic causation in the shallow mantle when the fossil Caledonian suture intersected the MAR. That event re-activated and melted a slab of Iapetus oceanic crust trapped in the suture from closure of Laurentia and Avalonia. They espouse that the GIR suggests in situ development along the buried orogenic plate boundary. In either case, effusive and voluminous magmatism constructed the elevated basalt plateau of Iceland that persists as the centerpiece of the NAIP.
ON THE LIGHTER SIDE
The streets and alleyways of Reykjavik are home to a thriving and creative art scene that brings color to otherwise dull parts of town and its often gray skies. Many are murals that adorn walls, underpasses and the sides of entire buildings. All whimsical, here are a few with geological inclinations.
|The real robotic source of Iceland's anomalous volcanism|
Sincere appreciation is extended to Gillian R. Foulger, PhD., British geologist, author of "Plates, Plumes and Paradigms" and Professor of Physics at Durham University, England for her patient emails that helped clarify issues regarding Iceland, the North Atlantic and mantle dynamics. As always, geologist, lecturer, author and Smithsonian Expert Wayne Ranney (here) provided invaluable geological assistance. I am also grateful to Sonia Didriksson, my local Icelandic connection for all things geographical and unpronounceable and for the excellent photographic contributions of Joel Feingold and renown Icelandic photographer Sigurður Stefnisson (here).
• A Cool Model for the Iceland Hotspot by G.R. Foulger and Don Anderson, Journal of Volcanology and Geothermal Research 141, 2005.
• An Alternative Model for Iceland and the North Atlantic Igneous Province by G. R. Foulger et al, On-line. No date provided.
• An Iceland Hotspot Saga by Ingi Þorleifur Bjarnason, Jökull 58, 2008.
• Earth Evolution and Dynamics – A Tribute to Kevin Burke by Trond Helge Torsvik et al, Canadian Journal of Earth Sciences, 2016.
• Fixity of the Iceland “Hotspot” on the Mid-Atlantic Ridge: Observational Evidence, Mechanisms, and Implications for Atlantic Volcanic Margins by Erik R. Lundin et al, GSA Special Paper 388, 2005.
• Frontiers in Large Igneous Province Research by Richard E. Ernst et al, Lithos 79, 2005.
• Iceland - Classic Geology in Europe 3 by Thor Thordarson and Ármann Höskuldsson, 2015.
• Iceland is Fertile: The Geochemistry of Icelandic Lavas Indicates Extensive Melting of Subducted Iapetus Crust in the Caledonian Suture by G. R. Foulger et al, Source On-line.
• Icelandic Glaciers by Helgi Björnsson and Finnur Pálsson, Jökull 58, 2008.
• Icelandic Rocks and Minerals by Kristján Sæmundsson and Einar Gunnlaugsson, 2002.
• Late Weichselian History of Relative Sea Level Changes in Iceland During a Collapse and Subsequent Retreat of Marine Based Ice Sheet by Halldór G. Pétursson et al, Cuadernos de Investigacion Geografica, 2015
• Mid-Ocean Ridge Jumps Associated with Hotspot Magmatism by Eric Mittelstaedt et al, Earth and Planetary Science Letters 206, 2008.
• NE Atlantic Break-up: A Re-examination of the Iceland Mantle Plume Model and the Atlantic–Arctic Linkage by E.R. Lundini and A.G. Gore, Petroleum Geology: North-West Europe and Global Perspectives - Proceedings of the 6th Petroleum Geology Conference, 2005.
• North Atlantic Igneous Province: A Review of Models for its Formation by Romain Meyer et al, On-line, No date provided.
• Plate Boundaries, Rifts and Transforms in Iceland by Páll Einarsson, Jökull 58, 2008.
• Postglacial Volcanism in Iceland by Thorvaldur Thordarson and Ármann Höskuldsson,
Jökull 58, 2008.
• The Glorious Geology of Iceland's Golden Circle by Agust Gudmundsson, 2017.
• The Greenland–Iceland–Faroe Ridge Complex by Arni Hjartarson et al, Geological Society of London, 2017.
• The “Plate” Model for the Genesis of Melting Anomalies by Gillian R. Foulger, GSA Special Paper 430, 2007.
• Volcanism in Iceland in Historical Time: Volcano Types, Eruption Styles and Eruptive History by T. Thordarson and G. Larsen, Journal of Geodynamics 43, 2007.