The fissure system of the Dyngjufjöll central volcano, with the Askja caldera, extends from the north margin of Vatnajökull some 100 km to the north. Volcanic activity in Dyngjufjöll is associated with two fissure systems, with slightly different strikes. These swarms are tangential to the east and west boarders of the present Askja caldera but intersect south of Dyngjufjöll.

The Askja lake is bounded in the south and east by precipitous cliffs and to the north and west by cliffs cutting the lavas that cover the bottom of the Askja caldera. The present level of the lake is 50 m below the bottom of the caldera. Soundings of the Askja lake reveal a roughly circular crater-like depression with a maximum depth of 224 m. Vigorous geothermal activity is manifest on the eastern and southern boarders of the lake and in wintertime small patches west of the lake remain ice free. A large crater, Víti, just northeast of the subsistence, contains fumaroles and a small pond with warm water.


In the following the geological history of the Dyngjufjöll complex is discussed, starting with the Pleistocene formation and ending with volcanic activity in the 19th and 20th centuries.

The Pleistocene formations in Dyngjufjöll are the oldest exposed rocks in the area. The age distribution of different formations between respective glaciations during the Pleistocene is, however, not known at present. Clear tillite horizons have not been identified and absolute dating is problematic due to the basaltic composition of the rocks. Recurring stratigraphic sequences, which can be associated with specific environmental factors, degree of alteration of volcanic glass and significant change in magma types erupted, allow a division of the Pleistocene formation into four units.

The oldest of these units is found on the flanks of the volcanic structure as isolated hillocks of hyaloclastite in an advanced state of alteration and partially buried by Postglacial lavas. These hyaloclastites were formed in shallow water with textures ranging from layered Surtseyjan tuffs to mass flow remnants, sometimes mixed with inclusions of laminated water lain sediments.

The southern part of the complex (Thorvaldsfjall), rising 800 m above the base, can be divided into three units, a basal and top units composed of near aphyric olivine tholeiite and a central plagiophyric unit. All units contain the same lithological facies consisting of pillow lavas, air fall Surtseyjan tuffs and mass flows. All three units were formed in a similar shallow water environment on erupting fissures parallel to the strike of the fissure system.

Distribution of the easily recognizable plagiobasalt indicates an erupting fissure close to a water level where some craters on the fissure were totally submerged and producing pillow lavas which flowed downhill into deeper water, while other craters on the same fissure produced fragmented material which is found both below and on the top of the pillow lava flows.

The similarity between the lithologies of the different units and the specific environment required to produce these lithologies constrains the period(s) of formation within a limited timeframe during deglaciation(s) when the volcanic complex had become ice-free but was surrounded by and partly submerged in marginal lakes dammed by the thinning glacier. It is conceivable that glacier unloading triggered a large but short lived production pulse leading to the formation of the units, each unit representing a separate deglaciation process.

The uppermost unit of the sequence was probably formed during the waning stages of the last glaciation 10-12 kA ago. Embedded in the uppermost air fall deposits of this unit is a rhyolitic ash layer which is only found on high ground within the Dyngjufjöll complex but reappears in coastal areas in northern and eastern Iceland directly on top of glacial deposits. The age of this layer is approximately 9800 years B.P. At this time the Dyngjufjöll complex was icefree but since the layer has not been found in the area between Dyngjufjöll and the coast it can be assumed that it fell on the top of a thinning glacier.

Postglacial volcanic activity in the Dyngjufjöll complex is associated with two fissure swarms tangential to the east and west boarders of the present Askja caldera. In addition large amount of lava has erupted in close vicinity to Dyngjufjöll fissure system producing many lava shields of different sizes. To the north of the main massif are Kollóttadyngja, Flatadyngja, Litladyngja and Svartadyngja, (dyngja in this context is synonymous with shield) and Trölladyngja to the south. Together, the lavas from Dyngjufjöll and from these various shields form the lava desert known as Ódádahraun.

The Postglacial history of the area can be pieced together by studying the distribution of tephra layers on the lava surfaces combined with aerial photographs and field studies of the lava stratigraphy.

Dated tephra layers in Ódádahraun

Layer "a" (Veidivötn) 1477 AD   
Öræfajökull  1362 AD   
Hekla 1 1158 AD   
Hekla 3  2900 BP   
Unknown origin ca. 3500 BP   
Hekla 4  4500 BP   

Tephra layers from Hekla and other volcanoes in Iceland provide useful time markers for the latter part of Holocene. The oldest Hekla layer, H5, which has been dated at 7-8000 years has, however, not been identified in the area.

Large amounts of lava were produced from the western branch of the fissure system across the Dyngjufjöll complex before the deposition of Hekla layer H4. These lavas flowed mainly towards NW and SW but also towards E into a small caldera, of early Postglacial age, situated north of the present Askja caldera. This caldera became filled with lavas from the western fissures and overflow occurred towards NE and NW.

Two shield volcanoes, Kollóttadyngja and Svartadyngja, were formed during this pre-H4 period. Kollóttadyngja is presently surrounded by younger lavas and the distribution of its lavas is therefore unknown. Svartadyngja is also surrounded by younger lavas except towards SE where lavas from this volcano can be traced south of Herðubreiðartögl and further east until they disappear beneath fluvial sediments from Jökulsá á Fjöllum.

The Askja caldera collapse dates back to this period. Postglacial age of the Askja caldera is indicated by the subaerial lavas filling the older caldera subsidence to the north which were subsequently cut by the younger Askja collapse. Due to topographic changes caused by the Askja subsidence, the older caldera was detached from vent areas which might otherwise have poured lavas onto its surface. Therefore, the Askja subsidence postdates the surface lavas in the older caldera. These surface lavas appear to be older than H4 so that the Askja subsidence occurred sometime between complete deglaciation (ca 6000-7000 years ago) and the deposition of H4, 4500 years ago. The formation of Askja must have occurred in the earlier part of this period since the time before H4 was sufficiently long to fill the Askja caldera to the Öskjuop threshold and subsequently to deposit lavas in a fan reaching past Herðubreiðartögl and to the river Jökulsá á Fjöllum.

Two lavas belonging to the Dyngjufjöll fissure system erupted in the area north of Dyngjufjöll during the period between the deposition of Hekla 4 and the "unknown" layer. In this period the lava shield Flatadyngja, SE of Kollóttadyngja, was formed. Lavas from this shield flowed to the north of Herðubreið and through the pass between Herðubreið and Herðubreiðartögl.

In the period between the "unknown" layer and Hekla 3 activity in the Dyngjufjöll volcanic system was concentrated within the Askja caldera. Lavas covered the caldera bottom and escaped through Öskjuop. These lavas did, however, not reach as far as the lavas older than Hekla 4. Shortly before the deposition of Hekla 3 the youngest lava shield in the area, Litladyngja, formed slightly to the west of the eastern branch of the Dyngjufjöll volcanic system, between Kollóttadyngja and Dyngjufjöll.

In the period between Hekla 3 and Hekla 1158 lavas may have issued inside the caldera and covered by later lavas. No lava flow has been found outside Öskjuop with Hekla 1158 as the oldest layer.

In the period between Hekla 1158 and Öræfajökull 1362 activity was limited to craters inside the Askja caldera. Lavas on the caldera floor from this period appear to have originated from both the eastern and western branch of the volcanic system.

In the period between Ö 1362 and layer "a" 1477 all the activity was within the caldera. It appears that a large part of the caldera floor was covered with lava in a single eruption from the western branch of the volcanic system. This eruption occurred shortly before 1477 AD, because the "a" layer is deposited immediately on top of the lava. This lava extends out through Öskjuop. Although the eruptions since the Hekla 1158 layer was deposited fall within historical time in Iceland, written records are not specific about volcanic activity in Dyngjufjöll.

A rough esimate of lava volumes produced in and around the Dyngjufjöll complex within the time periods defined by the tephra layers allows an order of magnitude determination of volcanic productivity in the area during Holocene (Sigvaldason et al. 1992). It appears that volcanic productivity in the period from the end of glacial rebound (ca. 8000 years B.P) until the deposition of Hekla tephra H4 at 4500 years B.P. was higher by a factor of 30 than the productivity since the deposition of Hekla tephra H3 2900 years B.P. This change in productivity has been explained as a melting anomaly in the mantle caused by pressure changes by loading and deloading of glacier ice, which in this central part of Iceland reaches a thickness of up to 2000 meters during peak glaciation.


The 1875 eruption

The first eruption in historical time which is well documented is the explosive eruption of 1875, which resulted in the formation of the nested caldera Öskjuvatn. At the time, however, Askja was practically unknown. Located within a large desert it had never been described. All early observations were made from a distance of about 100 km until shortly before the plinian eruption when a group of young farmers from the Mývatn district visited Askja on February 15, 1875 and gave a short description of what was happening.

Precursors to the eruption of 1875 were reported as early as February 1874. Unusually dense steamclouds were observed in Dyngjufjöll from afar. During the two last weeks of December 1874 strong and frequent earthquakes were felt in North Iceland. On 1 January 1875 a column of smoke rose in a southerly direction from the Mývatn district and on the following day frequent earthquakes were felt in that area, but cloudy weather resulted in poor visibility soon afterwards. Minor ash fall was observed in Northern Iceland. Earthquake activity came to an end a little later.

On 18 February 1875 a basaltic fissure eruption started in Sveinagjá, 40 to 70 km north of Askja, on the continuation of the Dyngjufjöll volcanic system. This eruption lasted, with short quiet intervals, for several months and produced 0.2-0.3 km3 of lava. It has been suggested that these lavas came from a basaltic holding chamber beneath the Askja caldera, travelling up to 70 km by lateral diking. This, however, is speculative and difficlt to constrain.

On 15 February 1875 a subsidence was observed in the SE corner of the caldera. An area of 3-4 hectares had subsided as much as 10 m. To the west of the subsidence there was an active crater, but the description indicates no production of lava or tephra, only violent boiling in mud pots at the time of observation. Pumice and ash had been deposited on the mountains to the northeast of the active area.

A plinian outbreak started in the early morning of 29 March 1875. Tephra fell at 3:30 a.m. in Eastern Iceland, 70 km to the east of Askja, carried by strong winds. The tephra was grayish in color, fine grained and so wet, that it could be formed like clay. The first phase lasted about one hour. At 5:30 in the morning the air had cleared somewhat, but shortly afterwards the main eruption started. Light-brown pumice, with ever increasing grain size, continued to be produced until noon. Although tephra-fall stopped in populated areas at noon, the eruption continued until the morning of the following day. The tephra contained a number of white and black fragments of volcanic glass, as well as basaltic sand, which was much heavier than the pumice and remained in place when everything else was removed by wind and water.

When visited in 1876 a triangular area of subsidence had formed in the southeast corner of the Askja caldera. It measured 4580 m in east-west direction and about 2500 m north-south. The deepest part lay 234 m below the bottom of the caldera. The bottom of the subsidence was occupied by a lake 1255 m in diameter. Around the lake were numerous concentric step-like faults which had cut the northern rim of the subsidence. At that time the lake level was 150 m below the caldera floor. The subsidence was practically filled with water to the present level in 1907.

The Askja lake is bounded in the south and east by precipitous hyaloclastite cliffs and to the north and west by cliffs cutting the lavas that cover the bottom of the Askja caldera. The present level of the lake is 50 m below the caldera floor. Bathymetric soundings of the Askja lake reveal a roughly circular crater-like depression with a maximum depth of 224 m. Vigorous geothermal activity is manifest on the eastern and southern boarders of the lake and in wintertime small patches west of the center of the lake remain ice free. A large crater, Víti, just northeast of the subsidence, contains fumaroles and a small pond with warm water. This crater was long believed to be the main vent of the 1875 eruption, despite early observations which clearly show that this was not the case.

The volume of the lake is 1.231 km3 and the area of the lake is 10.7 km2. The lake surface lies 50 m below the Askja caldera floor and the minimum volume above the lake surface is 0.54 km3. However, the subsidence cuts the mountainsides to the east and south of the lake, which are up to 400 m higher than the Askja caldera bottom. It is therefore difficult to give an exact estimate of the total subsidence. The absolute minimum value is 1.77 km3, but 2-2.5 km3 is probably a more realistic figure. The volume of freshly fallen tephra is 1.8 km3 or 0.43 km3 calculated as solid rock. The volume of the Sveinagjá basaltic lava is 0.2 to 0.3 km3. A dyke going from Askja to Sveinagjá 1 km high and 4 m wide has a volume of 0.16 km3. Adding these figures together gives a total volume of known and implied material to leave a magma chamber beneath Askja of 0.89 km3. This is less than half the volume of the subsidence.

The first phase of the 1875 Askja eruption was the production of basaltic tephra on January 1. This tephra is found inside the caldera but also on the outer flank of the caldera rim where it is sandwiched between layers of snow which fell during the winter 1874-1875. Within the Askja caldera the basaltic tephra is covered by thick layers of phreatic mud deposits which in turn are overlain by the products of the main silicic eruption of March 29.

The position of the main vents and the succession of events can be pieced together by studying the distribution of the eruption products. The siicic products can be divided into groups according to their physical appearance:

  1. Tephra ranging in colour and grain size from nonvesicular greyish-white, fine-grained tephra to vesicular pumice of light-brown to black color and with increasing grain size upward in the pumice deposit.
  2. Welded agglutinate, reddish to black in colour.
  3. Highly vesicular black lava.
  4. Less vesicular dikes of the same chemical composition as the lava and the tephra.
  5. Chunks of obsidian of various color shades from black to greenish, often partly vesiculated.
  6. Fragments of silicic rocks ranging from granophyric nodules to flint-like angular glasses.
Of the presently visible silicic eruption products a lava flow in the northeast corner of the subsidence was the first to appear on the surface. A dike observed along the eastern caldera fault was probably the feeder to that lava flow. This indicates that magma was first intruded along faults along the eastern boundary of the main Askja caldera. The time of formation of this lava is after the basaltic eruption on Jan. 1st and towards the end of phreatic activity since the lava is covered with a thin layer of phreatic mud below some 9 meters of silicic pumice from the subsequent plinian phase of the eruption.

Distribution of airfall tephra around the Askja lake has thickness maxima indicating proximity to producing craters which where arranged along a fissure extending across the present lake from ENE to SWS

Craters on this fissure erupted the bulk of the tephra produced in the plinian eruption. The first product was the nonvesicular grayish-white, fine-grained tephra. The prevailing strong westerly wind blew the material to the east. A layer of fine-grained irregularly bedded tephra was deposited on the east side of the vent or vents, but it thins rapidly further east. On the windward side of the subsidence in the southwestern corner of the Askja caldera, the tephra layer is maximally 3 m thick, but thins out to zero within a few hundreds meters from the 1875 subsidence. A sharp discontinuity in the tephra layer between the fine-grained tephra and the overlying brown to black tephra represents the lull in activity from about 4:30 to 5:30 in the morning of 29 March 1875. This highly vesicular pumice was ejected with great vigor, and a larger proportion of the material was carried outside the caldera than during the earlier phase. The distribution of the material within the caldera is very distinct. A line drawn through the thickness maxima for the two phases indicates a parallel shift of the eruption fissure towards southeast.

A pronounced difference is between the material deposited to the east of the subsidence and that deposited on the windward side to the west. In the southwestern corner of the caldera the material, deposited directly on the grayish white tephra from the first phase, is a welded agglutinate without any pumiceous material, neither at the top nor the bottom of the layer. The welded agglutinate, erupted from the southernmost vent on the fissure, was deposited on the sloping wall of the subsidence and flowed back into the depression.

Towards the end of the eruption, vents located close to the northeastern end of the fissure produced material hot enough to weld into an agglutinate. This is covered by a layer of large pumice clasts, representing the waning phase of the eruption.

In summary the sequence of events was as follows:

  1. Injection of basaltic magma to shallow depth beneath the Askja caldera as a result of a rifting event affecting a large part of the 100 km long Dyngjufjöll fissure system. The rifting caused intense seismic activity and opening of fissures in the northern part of the system. Beneath Askja the basaltic intrusions were quenched by a large ground water body confined within the structural boundaries of the caldera with a resulting increase in geothermal activity as observed already in the beginning of 1874. Boiling of the ground water body continued for 13 months gradually exhausting the water reserve.
  2. By January 1st 1875 continued basaltic injection finally broke through the ground water barrier resulting in a short lived eruption of Surtseyjan tephra.
  3. On February 15th a significant subsidence was observed et the location of the present Askja lake. This subsidence was caused by the depletion of the water reservoir, a phenomena commonly observed in heavily exploited geothermal areas.
  4. On February 18 basaltic eruptions started on the fissure system 40 km to the North of Askja.
  5. The basltic intrusions beneath Askja engulfed a small body of silicic magma without intruding it. A slight heating of the silicic magma (not detectable in mineral equilibria) upset the volatile balance and initiated partial degassing and shift in oxygen fugacities of the magma which is displayed in the color stratification of the tephra..
  6. With the water barrier removed the silicic magma was mobilised first by injection of dikes along the eastern caldera wall, which served as feeders to a dacitic lava flow.
  7. This led to pressure release and explosive boiling of the silicic magma with opening of a ENE-WSW striking fissure(s) and production of fine-grained nonvesicular tephra for about one hour in a plinian eruption. After this initial phase the intensity of the eruption was somewhat reduced for about one hour before the beginning of a main plinian phase, which lasted for about 6 hours.
  8. The phreatic explosion crater Víti was formed after the plinian eruption came to an end and phreatic activity continued for several months or years after the eruption.
Volcanic eruptions in the 20th century

In the period between 1920 and 1930 four basaltic eruption occurrred in and around the Askja lake subsidence producing small lava fields and building an island in the Askja lake. A fifth eruption occured on the southern flank of the Dyngjufjöll complex during this period.

The latest volcanic event in Dyngjufjöll occurred in 1961

Chemical analysis of samples from the northern rift zone. NAL 7: Herðubreið, NAL 8: Vaðalda, NAL 18: Upptyppingar, NAL 31: Kollóttadyngja, NAL 3: Flatadyngja (?), NAL 10: Lindakeilir.
Rock no. 
NAL 18
NAL 31
NAL 10
SiO2  48.53 48.83  47.86 49.22 49.80  48.17 
Al2O3 16.30 16.14  15.36 15.37 15.25  13.62
TiO2 1.71 0.70  1.68 1.74 2.03  3.35
Fe2O3 1.77 1.85  2.27 3.50 2.52  3.61
FeO 9.01 7.68  9.27 7.54 9.54  11.09
MnO 0.17 0.15  0.17 0.16 0.18  0.22
MgO 6.85 8.46  8.15 6.85 6.70  5.43
CaO 12.51 13.93  11.48 12.41 12.62  10.09
Na2O 2.38 1.95  2.20 2.42 2.44  2.83
K2O 0.23 0.07  0.36 0.38 0.25  0.65
P2O5 0.12 0.05  0.12 0.11 0.12  0.12
H2O 0.33 0.19  0.36 0.31 0.56  0.85


Table. The Dyngjufjöll rock suite. Víti: 1875 tephra, B: Bátshraun 1923 lava, M: Mývetningahraun 1926 lava, 797: Xenolith 1875 eruption, NAL 15: Pleistocene basalt.
Rock no.
NAL 15
SiO2 72.10 50.87  50.44 77.81 48.77 
Al2O3 13.35 12.92  13.07 12.48 15.34 
TiO2 0.85 2.95  2.72 0.21 1.56 
Fe2O3 2.85 4.67  1.38 2.13  
FeO 3.69 13.19  11.40 0.48 8.74 
MnO 0.12 0.26  0.22 0.02 0.16 
MgO 0.96 5.40  4.93 0.10 9.56 
CaO 2.19 9.02  8.58 1.13 12.00 
Na2O 3.55 2.46  2.40 3.00 2.21 
K2O 2.39 0.76  0.69 2.59 0.24 
P2O5 0.19 0.26  0.14    
H2O 0.17 0.69  0.76 0.34  


Brandsdóttir, B.: Historical accounts of earthquakes associated with eruptive activity in the Askja volcanic system. Jökull, v. 42, p. 1-12, 1992.

Brown, G.C., Everett, S.P., Rymer, H., Garvie, D.W., and I. Foster: New Light on caldera evolution - Askja, Iceland. Geology, v. 19, p.352-355, 1991.

Camitz, J., Sigmundsson, F., Foulger, G., Jahn, C.-H., Völksen, C. and P. Einarsson: Plate boundary deformation and continuing deflation of the Askja volcano, North Iceland, determined with GPS, 1987-1993. Bull. Volc. V. 57, p. 136-145, 1995.

Helgason, Ö., Óskarsson, N. and G.E. Sigvaldason: Oxygen fugacity stratification of a magma chamber revealed by Mössbauer spectroscopy: Evidence from the 1875 Askja eruption, N. Iceland. Hyperfine Interactions, v. 70, p. 989-992, 1992.

Macdonald, R., Sparks, R. S. J., Sigurdsson, H., Mattey, D. P., McGarvie, D. W. and R.L. Smith: The 1875 eruption of Askja volcanoc Iceland: combined fractional crystallization and selective contamination in the generation of rhyolitic magma. Min. Mag. V. 51, p. 183-202, 1987.

Moore, J.G. and Lewis C. Calk: Degassing and differentiation in subglacial volcanoes, Iceland. J. Volc. Geotherm. Res., v. 46, p. 157-180, 1991.

Sigurdsson, H. and R.S.J. Sparks: Petrology of rhyolite and mixed magma ejecta from the 1875 eruption of Askja, Iceland. J. Petrol. v. 22, p. 41-84, 1981.

Sigvaldason, G.E.: Structure and products of subaquatic volcanoes in Iceland. Contr. Mineral and Petrol., v. 18, p. 1-16, 1968.

Sigvaldason, G.E., Annertz, K. and Magnus Nilsson: Effect of glacier loading/deloding on volcanism: postglacial volcanic production rate of the Dyngjufjöll area, central Iceland. Bull Volcanol., v. 54, p. 385-392, 1992.

Sparks, S.R.J. and Sigurðsson, H., 1977. Magma mixing: a mechanism for triggering acid explosive eruptions. Nature, 267, 315-318.

Sparks, R. S. J. and J.V. Wright: Welded air fall tuffs. Geol. Soc. Am., Special Paper 180, p. 155-166, 1979.

Sparks, R .S. J., Wilson, L. and H. Sigurðsson: The pyroclastic deposits of the 1875 eruption of Askja, Iceland. Phil. Trans. Royal Soc. London, v. 29, p. 241-273, 1981.

Thorarinsson, S. and Sigvaldason, G.E., 1962. The eruption of Askja, 1961. A preliminary report. Amer. J. Sci., 260, 641-651.

Tryggvason, E.: Ground deformation in Askja, Iceland: its source and possible relation to flow of the mantle plume. Journal of Volcanology and Geothermal Research, 39 (1989) 61-71.

Werner, R., Schmincke, H.-U. and G.E. Sigvaldason: A new model for the evolution of table mountains: volcanological and petrological evidence from Herðubreið and Herðubreiðartögl volcanoes (Iceland). Geol. Rundsch., v. 85, p. 390-397, 1996.