Icelandic Glaciers

Home , Climate of Iceland, Subglacial eruption

Reviewed research article

Helgi Björnsson and Finnur Pálsson Institute of Earth Sciences, University of Iceland, Sturlugata 7, 101 Reykjavík, Iceland [email protected] [email protected]

Abstract – Some 11% of Iceland is covered by glaciers. They contain 3,600 km3 of water, equivalent to a 35-m-thick ice layer spread evenly over the whole country; if melted, it would raise global sea level by 1 cm. This is Iceland’s greatest water storage, corresponding to the precipitation of 20 years. Dynamic in nature, these glaciers are responsive to climate fluctuations and affect their environment profoundly. Also, they lie over active volcanoes; these induce jökulhlaups that can threaten areas of habitation. The country’s glaciers feed its largest rivers and currently provide at least one-third of its total runoff. Since a general glacier recession set in at the end of the 19th century, the largest icecap, Vatnajökull, has decreased by about 10% in volume (300 km3), contributing 1 mm to the concurrent rise in sea level. During the last ten years, ice losses have accelerated, thereby detracting 2.7% (84 km3) from the total icecap volume. Typically, radiation provides two-thirds of the melt energy, turbulent fluxes one-third. However, transitory volcanic eruptions and continuous geothermal activity at the bed of Vatnajökull added some 5.5 km3 to surface melting during the 1990s, with one particular volcanic eruption melting 4.0 km3. In all of Iceland’s major icecaps, surges account for a significant portion of total mass transport through the principal outlet glaciers, playing an important role in outlet dynamics and hydrology. Taking the 20th century as a whole, surges contributed at least 10% to the total ice transport to ablation areas of Vatnajökull. Plausible future climate scenarios, coupled with models of mass balance and ice dynamics, suggest that the main icecaps will lose 25% to 35% of their present volume within half a century, leaving only small glaciers on the highest peaks after 150–200 years. Glacier meltwater runoff will peak after 50 years, then decline to present-day values by 100 years from now. When the glaciers have disappeared, the entire river discharge will come directly from precipitation.

INTRODUCTION 550–600 m. Heavy snowfall is frequently induced by An island of 103,000 km2, Iceland lies in the North cyclones crossing the North Atlantic, where air and Atlantic Ocean, close to the Arctic Circle. Thanks water masses of tropical and arctic origins meet. At to the warm Irminger Current, the land enjoys a rel- higher elevations, this leads to snow accumulation. atively mild oceanic climate and small seasonal vari- At present, about 11% of the country is covered by ations in temperature. Average winter temperatures glaciers (Björnsson, 1978, 1979; Figure 1). hover around 0◦C near the southern coast, where the Classiﬁed as "warm-based" or "temperate", Ice- average temperature of the warmest month is only landic glaciers are dynamic in nature. Not only do 11◦C and the mean annual temperature is about 5◦C they respond actively to climatic ﬂuctuations, but con- (Einarsson, 1984). Along the northern coast, the cli- stitute long-lasting reservoirs of ice that turns to melt- mate is affected by the polar East Greenland Cur- water and feeds the country’s main rivers, some of rent, which occasionally brings sea ice. In the central which have been harnessed for hydropower. These highlands, permafrost can be found at altitudes above icecaps conceal unexplored landforms and geologi-

JÖKULL No. 58, 2008 365 H. Björnsson and F. Pálsson

Figure 1. Topography of Iceland, with glacier distribution. The main icecaps are bordered by smaller glaciers. The inserted geological map shows the active volcanic zone and the central volcanoes. – Íslandskort sem sýnir legu helstu jökla. cal structures, including active volcanoes, geother- by glacial and fluvioglacial sediments in late glacial mal sites and subglacial lakes. Catastrophic floods and early Holocene periods. In addition, the topogra- (jökulhlaups) from geothermal and volcanic locations phy and sediments of near-shore marine environments are frequent. These floods have periodically threat- have been heavily influenced by glacial erosion and ened inhabited regions, damaged vegetation, dis- deposition. The impact of glacial rivers is evidenced rupted roads and communications, and even temporar- by deeply eroded canyons and sediments transported ily deterred fish from entering coastal waters. About onto sandur deltas. Iceland’s specially-named Pala- 60% of today’s glacial area is underlain by active gonite Formation is largely the product of subglacial volcanoes, occasioning intensive studies of glacier- volcanic activity that was later subjected to erosion. volcano interaction. Because of their profound environmental effects, During Pleistocene and post-glacial times, the is- the Icelandic glaciers and associated phenomena have land and its surrounding sea-floor topography have long attracted interest, so that a wealth of general in- been significantly shaped by glacial erosion and formation is available on the country’s volcanic and glacial or fluvioglacial deposits. Glaciers have carved hydrological events. More recently, a considerable alpine landscapes characterised by cirques, sharp amount of more precise hydrological, glaciological mountain peaks, broad lowlands, and long, steep U- and volcanological data have been collected, while shaped valleys or narrow fjords. The largest agri- meteorological observations on the glaciers have re- cultural regions in the south and west were created vealed relationships between climate and mass bal-

366 JÖKULL No. 58, 2008 Icelandic Glaciers ance. The accumulated glaciological data now allow high surface albedo. Icelandic higher-altitude sum- for a modelling of the mass balance and ice dynam- mers are chilly in any case, so that in the uppermost ics and for using coupled models in order to evalu- parts of the glaciers, days when melting occurs num- ate glacier response and glacial runoff in conjunction ber only 10 to 20 per year, though at lower levels the with any given past or future climate change. This ablation season typically lasts three to four months paper reviews the data that have been gathered and (June through mid-September). modelling work that has been carried out. The southernmost outlets of Vatnajökull and Mýr- dalsjökull, not far removed from the sea coast, de- GLACIER DISTRIBUTION AS RELATED scend to 100 m elevation or less (lowest 20 m), where TO CLIMATIC AND TOPOGRAPHICAL even winter balances are slightly negative. Even CONDITIONS though annual precipitation there may total up to 1,500 mm or more, most of it falls as liquid water, re- The regional distribution of glaciers in Iceland in- sulting in net ablation during every season of the year. dicates how precipitation arrives with prevailing Summertime net losses typically measure 9 m (water southerly winds (Figure 1). On the highest south- equivalent) at the snouts of Vatnajökull (terminating at ern slopes of Vatnajökull and Mýrdalsjökull, i.e. elevations around 100 m). For glacier outlets in cen- above 1,300 m, annual precipitation exceeds 4,000– tral Iceland, which terminate at 600–800 m, summer 5,000 mm, peaking at 7,000 mm (Figure 2), while balances at the snouts range from -4 to -6 m. it reaches 3,500 mm on Hofsjökull and Langjökull. Also noteworthy is that the largest icecaps are situated Central Iceland also has several steep mountain in the southern and central highlands (Table 1). On peaks reaching over 1,400 m above sea level and top of the larger icecaps, average temperatures are be- maintaining small glaciers. The dry inland regions low or close to freezing throughout the year, with most farther north receive an annual precipitation of only of the precipitation falling as snow. While the sum- 400–700 mm, lifting the glaciation limit even higher mer balance is normally negative in the central portion than 1,600 m in the rain shadow north of Vatna- of Vatnajökull, it may turn to slightly positive (up to jökull. In northern coastal areas this limit descends 0.5 m) when repeated cold spells with northerly winds to 1,100 m, as evidenced by over 100 small cirque bring fresh snow to the glacier, thereby maintaining a glaciers and corries, frequently facing north and lo-

Figure 2. Mean annual precipitation of Iceland in 1961–1990 (Crochet et al., 2007). – Meðalársúrkoma á Íslandi 1961– 1990.

JÖKULL No. 58, 2008 367 H. Björnsson and F. Pálsson

Table 1. Glaciers in Iceland. General features. – Skrá um tölur sem lýsa jöklum: ﬂatarmáli, rúmmáli, hæð yﬁrborðs, botns og jafnvægislínu, afkomu og skriðhraða. Glacier Area Volume Mean Surface elev., Bed elev., Max. ice (year) thickness range and mean range and mean thickness km2 km3 m m a.s.l. m a.s.l. m m a.s.l. Vatnajökull 8,100 3,100 380 0–2110; 1215 -300–2110; 670 ∼950 (∼2000) Langjökull 900 190 210 430–1440; 1080 410–1330; 865 ∼650 (∼2000) Hofsjökull 890 200 225 620–1790; 1245 470–1650; 1020 ∼760 (∼2000) Mýrdalsjökull 590 140 240 120–1510; 995 15–1375; 760 ∼750 (∼2000) Drangajökull 160 ∼24 ∼150 60–910; 665 ∼260 (∼1990) Eyjafjallajökull 80 190–1635; 1135 (∼1990) Tungnafellsjökull 43 920–1520; 1335 (∼1990) Þrándarjökull 27 760–1235; 1035 (∼1945) Eiríksjökull 24 550–1665; 1390 (∼1990) Þórisjökull 22 770–1325; 1165 (∼1990) Tindfjallajökull 20 650-1450; 1190 (∼1990) Snæfellsjökull 16 0.5 30 600–1440; 980 600–1440; 950 ∼110 (∼2000) Torfajökull 13 780–1170;1015 (∼1990) Hrútfell 8 690–1400; 1205 (∼1990) Hofsjökull in Lón 6 870–1145; 1045 (∼1990) Gljúfurárjökull 3 610–1355; 1015 250 (∼1990) Tröllaskagi ∼190 glaciers Total ∼11,100 ∼3,600

Table 1., cont. Glacier Winter bal. Spec. runoff Winter bal., Summer bal., Aver. ann. net bal. Typ. surf. vel. at zero net mass bal. range range mw.eq. ls−1km−2 m w.eq. m w.eq. m w.eq. m a−1 Vatnajökull 1.6 50 -4.5 to 6.5* -9.5 to 2.5* -0.45* 25–350 Langjökull 1.8 55 0 to 4.3** -8.5 to -0.5** -1.3** 20–60 Hofsjökull 1.8 60 -.2 to 4.3***-6.8 to 1.31*** -0.53*** 20–60 * Measured in 1991–92 to 2006–07 ** Measured in 1996–97 to 2007–08 *** Measured in 1987–88 to 2005–06

368 JÖKULL No. 58, 2008 Icelandic Glaciers

cated above the main valleys and around the high- larger icecaps subsist thanks mainly to their own est peaks, those reaching altitudes of 1,300–1,500 m. height. Even though Vatnajökull, for instance, lies These summits receive an annual precipitation of up on a highland plateau of 600–800 m above sea level, to 2,000 mm, brought to a large extent by northerly with 88% of its bed lying above 600 m, it is actually winds. Accumulation is locally increased through only 20% of the bed that exceeds 1,100 m, which is snow drift, whereas melting is in many places reduced the glaciation limit for southern Iceland. Six moun- by the shadowing effect of narrow valleys. tain ranges form the main glaciation centres within In the extensive northwestern peninsula called the Vatnajökull, peaking at 1,200 to 2,000 m: Gríms- West Fjords, annual precipitation reaches 3,000 mm, fjall (1,700 m), Bárðarbunga (1,800 m), Kverkfjöll and the glaciation limit registers lowest in Iceland, at (1,930 m), Öræfajökull (2,000 m), Esjufjöll (1,600) 600–700 m. The mountain plateau on this peninsula and Breiðabunga (1,200 m). Another example of how lies between 700 m and 900 m above sea level, so an icecap has little other than ice reaching above the that a number of niches and some 10 small cirque glaciation limit is Mýrdalsjökull, which is underlain glaciers are found at elevations between 600 and by a huge central volcano that has rims of 1,300– 700 m. The highest part of the peninsula is covered 1,380 m around a caldera 650–750 m deep. This by the icecap Drangajökull, Iceland’s northernmost means that only 10% of the Mýrdalsjökull bed rises glacier. It terminates below 200 m and, together with over the glaciation limit of 1,100 m above sea level. the Breiðamerkurjökull outlet of Vatnajökull, extends Hofsjökull also covers a major central volcano, with closest to the sea of any of the country’s glaciers. rims of 1,300–1,650 m surrounding a caldera that drops to an elevation of about 980 m. Around 20% of GLACIER GEOMETRY the Hofsjökull bed exceeds an elevation of 1,200 m, while 11% surpasses 1,300 m. The Langjökull icecap Maps of the surface and subglacial topography of all covers a 50-km-long mountain chain that rises to only the major icecaps have been produced by interpolat- 1,000–1,250 m, placing a mere 5% of the bed above ing continuous profiles spaced about 1 km apart, us- 1,200 m. ing radio echo-sounding for ice thickness along with precision altimetry (Sverrisson et al., 1980; Björns- son, 1982, 1986a,b, 1988, 1996; Björnsson et al., MASS BALANCE AND MELTWATER 2000; Björnsson and Einarsson, 1990; Magnússon et DRAINAGE al., 2004, 2005a,b,c, 2007). Specific drainage basins Surface maps indicate the directions of large-scale ice have been delineated, glacier mass balance monitored, flow and show the limits of ice-drainage basins for and the meltwater contribution to various rivers esti- the principal rivers flowing from the glaciers (Fig- mated. Statistics derived from available maps are pre- ure 1). Since about 1990, annual mass balance mea- sented in Table 1 and Figure 3. Glacier surface and surements have been conducted at the largest ice- bedrock maps reveal previously undiscovered land- caps: Hofsjökull (since 1987/88), Vatnajökull (since forms and geological structures within the active vol- 1991/92), Langjökull (since 1996/97) and Dranga- canic regions and identify the locations of calderas, jökull (since 2004/05), and some measurements also volcanic centres and fissure swarms. Furthermore, exist for smaller icecaps (Björnsson, 1971; Björnsson maps of glacier surfaces are available, providing im- et al., 1998, 2002; Sigurðsson and Sigurðsson, 1998; portant details about surface elevations, structures and Sigurðsson et al., 2004, 2007). The mass balance data slopes, as well as charts of the various glacier out- provide a basis for estimating the meltwater contribu- lets. The geometry of subglacial meltwater cupolas tion to glacial river systems. Bedrock maps are used and of ice-dammed lakes in geothermal areas can also in conjunction with the glacier surface maps for delin- be viewed on present-day maps. eating the locations of water-drainage basins feeding Typically, only 10–20% of the bed of the glaciers glacial rivers (Figure 4). lies above today’s glaciation limit. Thus, Iceland’s

JÖKULL No. 58, 2008 369 H. Björnsson and F. Pálsson

Figure 3. Area and volume distributions, along with elevation, for Iceland’s largest icecaps: Vatnajökull, Langjökull, Hofsjök- ull and Mýrdalsjökull. Bedrock areas are shown with black and grey lines, glacier surfaces with blue lines. – Dreifing flatarmáls og rúmmáls jökla með hæð. Botn er dreginn með svörtum og gráum línum en yfirborð með bláum.

The average mass balance of Vatnajökull for glacier years 1991/92 to 2005/06 is shown in Figure 5, and the temporal variation in Figure 6. The winter balance was generally highest in the early 1990s, diminished to a minimum in 1996–1997, rose to a maximum in 2003, and since then has slowly de- clined. Summers were comparatively cold during the ﬁrst half of the 1990s, as reﬂected in low summer ablation. The high summer melting of 2000, on the other hand, was pri- marily attributable to warm, windy weather. On Vatnajökull, the annual net balance remained positive from 1991/92 to 1993/94, approached zero in 1994/1995, and has been negative since then (Figure 6). Vatnajök- ull has lost about 0.8 m a−1 since 1995/96, as an average over the whole glacier. The total mass loss of Vatnajökull in 1994/95 through 2005/06 was 9.2 m (water equivalent)or84km3 (which amounts to 6 times the average winter balance), and the icecap lost about 2.7% of its total mass. In addition to surface melting, continuous geothermal activity at the bed of Vatnajökull and transitory volcanic eruptions melted about 0.55 km3 a−1 on average in the 1990s, which equals only 4% of the total surface ablation of ∼13 km3 a−1 during one average year of zero mass balance. The volcanic eruption in Gjálp in October 1996 by itself melted ∼4.0 km3 of ice.

370 JÖKULL No. 58, 2008 Icelandic Glaciers

Figure 4. Water drainage basins at the glacier bed for the principal rivers draining Vatnajökull. Subglacial water-ﬁlled cupolas, located under depressions in the glacier surface, collect meltwater and periodically drain by means of jökulhlaups. In order to determine water divide locations on a regional scale, a model was employed which assumed that water pressure at the glacier bed is approximately equal to the ice overburden pressure. – Vatnasvæði helstu jökulﬂjóta sem falla frá Vatnajökli.

In years of zero mass balance, 55–65% of the early 1990s. Rain on the glacier during the five sum- glacier surface typically lies above the equilibrium mer months may add 10–20 l s−1 km−2 to specific line altitude. During the period of 1992 to 2007, the discharge from the glacier. accumulation area of the ice flow basins of Vatnajök- The mass balance records for Hofsjökull and ull varied from 20 to 70% of their total area (Figure Langjökull show characteristics similar to Vatnajök- 7), the equilibrium line altitude (ELA) fluctuated by ull. The net balance of Langjökull has remained 200–300 m (300 to 400 m), and the annual net bal- negative throughout the survey period of 1996/97 to ance ranged from plus to minus one metre. A 100-m 2005/06, while the accumulation area has varied from change in ELA affects the net mass balance of Vatna- 10% to 40% of the glacier. Total mass loss over the jökull by ∼0.7 m a−1. period 1996/97 to 2005/06 was 12.8 m (13.1 km3 ice), During years of zero net balance, the runoff from or 7% of the ice mass. In instances of zero mass bal- Vatnajökull relating specifically to the summer balance, annual turnover rates approximate 0.4% of total ance was about 50 l s−1 km−2, averaged over the en- volume of Vatnajökull and 0.8% of Langjökull and tire glacier and entire year, but dropped to half of this Hofsjökull. in years of the most positive mass balance, i.e. in the

JÖKULL No. 58, 2008 371 H. Björnsson and F. Pálsson

Figure 5. Maps of average mass balance (m water equivalent) for Vatnajökull, glacier years 1991/92 to 2005/06. Specific mass balance in m water equivalent, bw: winter balance, bs: summer balance, bn:annual net balance (see Björnsson et al., 1998). Mass balance was calculated by a stratigraphic method, based on measured changes in thickness and density relative to the summer surface at about 50 sites. – Kort af meðalafkomu á Vatnajökli frá 1991/92 til 2005/06; bw: vetrarafkoma, bs: sumarafkoma, bn:ársafkomaí metrum vatns. GLACIO-METEOROLOGY During a 100-day period every summer, several auto- matic meteorological stations have been operated on some of the Icelandic icecaps; these operations began on Vatnajökull in 1994 and on Langjökull in 2001. Radiation components have been measured directly in situ, whereas turbulent fluxes have been calculated corresponding to wind, air temperature and humidity in the boundary layer. As a rule, net radiation is shown to be the outstanding factor in melting, although it is occasionally equalled by eddy fluxes (Figure 8). During the melting period, radiation typically provides two-thirds of the melt energy, and turbulent fluxes one-third (Björnsson, 1972; Björnsson et al., 2005; Oerlemans et al., 1999; Guðmundsson et al., 2006). At higher stations on the icecap, turbulent exchange becomes less significant. There are sporadic cases of radiation contributing somewhat to melting when eddy fluxes are negative. Solar radiation absorption increases substantially when winter snow has disappeared from the ablation zones, exposing ice covered with tephra lay- ers; under these conditions albedo may decrease to as low as 10%. Net radiation comes to a peak in ablation areas in June, in accumulation areas in August. Turbulent fluxes increase during the summer, peaking in August and September. The atmospheric boundary layer is found to be dominated by katabatic flows, especially in the lower, steeper regions of each icecap. It is only during the passage of intense storms that the katabatic winds of ablation zones become negligible.

372 JÖKULL No. 58, 2008 Icelandic Glaciers

Figure 6. Temporal variations in Vatnajökull mass balance (for glaciological years 1991/92 to 2005/06). Spe- ciﬁc mass balance is given in m water equivalent, where bw stands for winter balance, bs summer balance and bn annual net balance. – Afkoma Vatnajökuls 1991/92 til 2005/06.

During volcanic eruptions, the impact of tephra GLACIER DYNAMICS fall is short-lived in the accumulation area, in most The average surface velocity has been estimated based cases only affecting ablation in the following sum- on summertime GPS measurements at most of the mer. The high ablation on Vatnajökull in the summer mass balance sites (Figure 9). In addition, velocity of 1997 was to a large extent caused by low average maps for extensive areas have been derived from satel- albedo following exposure of the tephra layer from lite data obtained from InSAR and SPOT (Björns- the Gjálp eruption in October 1996. Moreover, dust son et al., 2001a; Fischer et al., 2003; Magnússon originating from the jökulhlaup deposits on Skeiðar- et al., 2005b; Berthier et al., 2006). Steeply sloping ársandur in November 1996 was later spread by wind glaciers, whether hard- or soft-bedded, seem to move over broad expanses of the icecap. with sufficient speed to keep in balance with annual mass balance. In contrast, surge-type glaciers, char- The observed daily melting rates have been suc- acterised by gently sloping surfaces (typically 1.6– ◦ cessfully simulated by energy balance calculations 4 ), move too slowly to maintain a balance in relation based on meteorological observations on each glacier. to their mass balance rate (Björnsson et al., 2003). However, temperatures measured on the glacier itself Surge intervals vary between glaciers, lasting from do not provide the most successful degree-day pre- several years up to a century; moreover, surge fre- dictions of ablation; instead, temperature data from quency is most often neither regular nor clearly re- outside the glacier provide more reliable predictions, lated to glacier size or mass balance. Altogether, 26 when projected onto the glacier using a constant wet surge-type glaciers have been identified in Iceland, 2 adiabatic lapse rate according to elevation. Air tem- ranging in size from 0.5 to 1,500 km (Figure 10). peratures in the low-albedo neighbourhood of the About 80 surge advances have been recorded, extend- glacier indicate daily variations in global radiation ingfromdozensofmetresupto10km(Björnssonet flux more precisely than the damped boundary layer al., 2003). temperatures above the melting icecap itself.

JÖKULL No. 58, 2008 373 H. Björnsson and F. Pálsson

For all of the major icecaps, surges account for surges to mass transport exceeds even this. Viewing a significant portion of total mass transport through the 20th century as a whole, surges were responsible the main outlet glaciers and have important implica- for at least 10% of the total ice flux to ablation areas. tions for outlet dynamics and hydrology (Þórarinsson, Surges increase river sediment loads (and hence 1969; Björnsson, 1998; Björnsson et al., 2003; Mag- sediment concentrations) substantially, especially in nússon et al., 2005a). They reduce ice-surface slopes, the finest grain sizes (Pálsson and Vigfússon, 1996; alter glacier hypsometry through mass transport, and Pálsson et al., 2000; Sigurðsson, 1998). For one to increase the area and roughness of the glacier surface. two years following surges in Vatnajökull, the sedi- In the wake of a surge, the resulting surface rough- ment concentration of affected outlet rivers generally ening and ice deposition at low elevations accelerates attains 7–10 kg m−3. These concentrations are com- surface melting due to solar radiation and turbulent parable to those during glacier outburst floods. During heat exchange; thus, runoff to glacial rivers increases. 2 the 1963–1964 surge of Brúarjökull on the north side During the 1990s, ∼3,000 km of Vatnajökull (38% of Vatnajökull, the river Jökulsá á Brú had an aver- of the icecap area) was affected by surges, which −3 3 age suspended sediment concentration of 6.5 kg m . transported about 40 km of ice from accumulation ar- During surges that typically last less than one year, eas to ablation areas. This amounted to approximately a denudation rate has been observed of 14 mm a−1, 25% of the total ice flux to ablation areas during this averaged over the entire glacier bed (see Björnsson, period. For some outlet glaciers, the contribution of 1979).

Figure 7. Relationship between net annual balance (bn), accumulation area ratio (AAR) and equilibrium line altitude (ELA) for Vatnajökull (in glaciological years 1991/92 to 2005/06). – Tengsl ársafkomu (bn), stærðar safnsvæðis af heildarﬂatarmáli jökuls (AAR) og hæðar jafnvægislínu (ELA) á Vatnajökli 1991/92 til 2005/06.

374 JÖKULL No. 58, 2008 Icelandic Glaciers

Figure 8a. Contribution of energy fluxes during the ablation season of 2004 to meltwater discharge (QA) from the outlet Brúarjökull (1,600 km2) in NE Vatnajökull. The data consisted of meteorological measurements at three sites on the glacier and mass balance measurements at 15 sites. 8b. Relative contribution of various energy fluxes to the total energy provided for melting during the ablation season: QR total radiation, QRs short wave, QR1 long wave, QH turbulent fluxes, QHd sensible heat, QHl latent heat. – Framlag orkustrauma til leysingar og afrennslis (QA) á Brúarjökli sumarið 2004. QR heildargeislun, QRs sólgeislun, QRl himin- og jarðgeislun, QH varmastraumar með lofti, QHd skynbær varmi, QHl dulvarmi. Hægri mynd sýnir hlutfall af heild.

JÖKULHLAUPS which melts the glacier ice above and thereby causes a depression in the glacier surface. The relatively Glacier-related floods, called jökulhlaups or outburst low basal pressure potential under such depressions floods, are particularly frequent in Iceland. Their causes meltwater to accumulate in cupolas which gain sources are of three types: geothermal fields that con- size until becoming unstable and bursting out from stantly melt the glacial ice above them so that melt- the glacier in a jökulhlaup. One area to which par- water accumulates and drains periodically, volcanic ticular attention has been devoted is the Grímsvötn eruptions which melt the ice into water that drains vicinity of Vatnajökull, where there has been geother- without delay towards the glacier margin, and pre- mal activity for centuries (Figures 4 and 11). Using carious ice dams which form lakes at glacier mar- Grímsvötn lake as a natural calorimeter, data on the gins (Þórarinsson, 1974, 1975; Rist, 1955, 1970, mass balance of its drainage basin allow for estimat- 1973, 1976, 1981, 1984; Tómasson, 1973, 1974, ing the heat output which produces the fluid phase at 1996; Tómasson and Pálsson, 1980; Björnsson, 1974, the subglacial geothermal area as 1.2 TW (or about 1975, 1976, 1992, 2002; Jóhannesson, 2002; Guð- 30% of the total heat) and the output which produces mundsson et al., 1995, 1997, 2003; Flowers et al., the vapour phase as 3.0 TW (or about 70%) (Björns- 2004). At present, Iceland has some fifteen of the last- son 1983, 1988; Björnsson et al., 1982; Björnsson and named type of jökulhlaup source, i.e. marginal ice- Kristmannsdóttir, 1983; Björnsson and Guðmunds- dammed lakes (Þórarinsson, 1939; Björnsson, 1976). son, 1993; Guðmundsson et al., 2002). Active volcanoes and hydrothermal systems under- lie 60% of the area of Icelandic icecaps. Jökulh- Jökulhlaups may profoundly alter landscapes, laups drain regularly from six subglacial geothermal devastate vegetation, and threaten lives as well as the areas (about 1 km3a−1 of water). Geothermal ac- roads, bridges and hydroelectric plants along glacier- tivity under glaciers reflects the interaction of water fed rivers. The effects of jökulhlaups on the landscape with magmatic intrusions, creating geothermal fluid appear in massively eroded canyons and in sediment

JÖKULL No. 58, 2008 375 H. Björnsson and F. Pálsson

Figure 9. Measured summer velocity at mass balance sites on Vatnajökull in 2006, using kinematic Global Positioning System measurements. – Mældur hraði á Vatnajökli 2006. deposits on outwash plains. While the high sediment transporting many times this amount of sediment onto loads during both surges and jökulhlaups play an im- the Mýrdalssandur outwash plain; such single-episode portant role in building up some of the sandur deltas, deposits suffice to shift the coastline several hundred the total sediment mass transported during a jökulh- metres seawards. The Mýrdalssandur jökulhlaups are laup lasting only two to three weeks may be five times Earth’s largest contemporary floods, rivalled only by greater and the basal area excavated much more con- floods associated with the end of the last glaciation centrated than during a surge. This applies to typ- 11,500 years ago. Looking at the central highlands, ical jökulhlaups from the subglacial lake Grímsvötn where jökulhlaups are topographically constrained, down to Skeiðarársandur, which at 1,000 km2 is Ice- the thereby more focused erosion has created spectac- land’s most widespread outwash plain. During a vol- ular canyons, e.g. Jökulsárgljúfur. Some of the major canic eruption, total sediment transport in the result- Pleistocene river canyons may also have been formed ing jökulhlaup over Skeiðarársandur may be another through such catastrophic floods of glacial origin. five times higher. Volcanic eruptions under Mýrdals- jökull, on the other hand, each result in a jökulhlaup

376 JÖKULL No. 58, 2008 Icelandic Glaciers

Figure 10. The geographic distribution of Iceland’s surging glaciers, of which 26 have been identiﬁed, ranging in area from 0.5 to 1,500 km2. About 80 surges have been recorded, with advances ranging from dozens of metres up to 10 km (Björnsson et al., 2003). – Þekkt framhlaup í meginjöklum landsins.

RECENT GLACIER VARIATIONS icecaps on the highest mountains, such as Öræfajökull (Eyþórsson, 1931, 1935; Ahlmann, 1937, 1939, 1940; During the Climatic Optimum 7,000 years ago, the Ahlmann and Thorarinsson, 1937a,b, 1938, 1939; Pleistocene ice still remaining over Iceland disap- Þórarinsson, 1943, 1964, 1966; Guðmundsson, 1997; peared almost entirely, presumably leaving only small

JÖKULL No. 58, 2008 377 H. Björnsson and F. Pálsson

Figure 11. Cross-section from Grímsvötn to Skeiðarársandur along the ﬂow path of jökulhlaups. Middle: Lake levels of Grímsvötn, 1930–2000, with the lake level climbing until a jökulhlaup results. Below: Typical hydrographs of jökulhlaups from Grímsvötn (1934, 1938, 1954, 1976, 1982, 1986 and 1996). – Snið frá Grímsvötnum niður á Skeiðarársand. Miðja: Vatnshæð Grímsvatna 1930– 2000. Neðst: Rennslisrit jökulhlaupa.

Eiríksson et al., 2000). From about 8,000 to 3,000 et al., 2004; Flowers et al., 2007, 2008). As the years ago, the climate of Iceland was considerably climate became colder and precipitation presumably warmer and drier than at present, with average tem- increased, glaciers edged downwards from the high- peratures believed to have been ∼2◦C higher than in est summits. Capable of reacting quickly, some the period 1920–1960 (Einarsson, 1963; Vinther et steep Alpine glaciers attained their post-Würm max- al., 2006). imum. The outermost moraines fronting Kvíárjökull and Svínafellsjökull (outlets from Öræfajökull) prob- During Neoglaciation, Icelandic glaciers have un- ably stem from this period. Still other glaciers ex- dergone two outstanding periods of expansion. The panded over the highland plateau, developing into the ﬁrst one occurred during the climatic deterioration present Icelandic icecaps. Vatnajökull merged into around 500 years B.C., which was at the onset of Sub- a single icecap from its beginnings in outlets from atlantic time (Bergþórsson, 1969; Dugmore, 1989; several glaciation centres (Öræfajökull, Grímsfjall, Stötter, 1991; Sharp and Dugmore, 1985; Stötter et Bárðarbunga, Kverkfjöll, Esjufjöll and Breiðabunga). al., 1999; Wastl et al., 2001; Kirkbridge and Dug- more, 2001, 2006; Schomacker et al., 2003; Black

378 JÖKULL No. 58, 2008 Icelandic Glaciers

From the beginning of the settlement of Iceland tle Ice Age (Björnsson, 1996; Björnsson et al., 2001b; (∼874 A.D.) up to the thirteenth century, the cli- Magnússon et al., 2007; Nick et al., 2007). mate resembled that of the later warm period from During the 1890s, however, a general recession 1920 to 1960, with average air temperatures proba- ◦ commenced, becoming quite rapid after 1930. On bly 3 to 4 C below those of the Climatic Optimum the other hand, cooler summers became the rule af- (Bergþórsson, 1969; Ogilvie, 1992; Ogilvie and Jóns- ter 1940 (Figure 13), so that glaciers in general re- son, 2001; Kirkbridge, 2002. Thus land which is now treated more slowly during the 1960s, and many steep largely hidden by Breiðamerkurjökull was vegetated glaciers even started advancing around 1970. Since and even occupied by several farms. Then the sec- 1985, the once more warmer climate has steadily led ond outstanding period of icecap expansion during to more widespread retreat, and every non-surging Neoglaciation set in, called the Little Ice Age and outlet glacier in Iceland has been retreating since 1995 destined to last from the Middle Ages till the close (Bárðarson, 1934; Eyþórsson, 1931, 1963, 1962– of the 19th century. During this period, some glacier 1966; Þórarinsson, 1943; Rist, 1967–1987; Jóhannes- outlets advanced around 10–15 kilometres, devastat- son, 1986; Sigurðsson, 1998, 2005; Jóhannesson and ing vegetation along with several farms. The firn line Sigurðsson, 1998; Sigurðsson et al., 2007; Hanna et in southern Iceland crept down from 1,100 to 700 m al., 2004). The rate of retreat has accelerated due to in the latter part of the Little Ice Age (Figure 12). high summer melt, but no long-term changes in pre- For steeper outlets, glacier advance culminated in the cipitation have been observed. Since 1890, the lead- 1750s; for broad lobes from the plateau icecaps, it ing Vatnajökull outlets have drawn back as far as 2–5 culminated between 1850 and 1890. km, and the icecap’s volume has decreased by about 300 km3 (∼10%), contributing 1 mm to the rise in global sea level. In the warm period since 1995, the ice surface elevation in ablation zones has decreased by dozens of metres, with margins retreating at rates up to ∼100 m a−1. The southern outlet glaciers of Vatnajökull have been particularly vulnerable to such warming, since many had carved down into soft sediments during their Little Ice Age advance and there- fore now have beds lying hundreds of metres below the elevation of the current terminus. In addition, frontal lakes form during their retreat, facilitating and speeding it up through the melting and calving of the floating termini. Because Iceland’s major rivers are Figure 12. Firnline variations in southern Iceland dur- glacial in origin, and have in many cases been har- ing the past millennium. Adapted from Þórarinsson nessed to generate hydropower, this recession has had (1974). – Hjarnmörk við sunnanverðan Vatnajökul. a noteworthy hydrological impact. During the second half of the 20th century, the glacial contribution to While advancing, glaciers excavated their sedi- the country’s runoff was estimated to be about 30% ment floor; for example, some of the most active (1,500 m3 s−1 of 5,000 m3 s−1) (Rist, 1956; Tómas- southern Vatnajökull termini typically excavated their son, 1981, 1982; Jónsdóttir, 2008). These estimates, beds down to hard surfaces 200–300 m below sea however, were for years before the increased summer level. One instance is the 20–km-long trench, 2–5 melt following 1995/96. Current glacier runoff com- km wide and extending to 300 m below sea level, prises at least one-third of total runoff. River courses that was created as Breiðamerkurjökull advanced and have also changed, leading to problems for farmers ploughed away sediment during the course of the Lit- and the Road Administration.

JÖKULL No. 58, 2008 379 H. Björnsson and F. Pálsson

Figure 13. Summer temperature and winter precipitation at several Icelandic meteorological stations in the 19th and 20th centuries. Time series are ﬁltered using the 11-year triangular running average. – Meðalhiti sumars og vetrarúrkoma á nokkrum veðurstöðvum.

FUTURE OUTLOOK

Numerical models have been developed for describing glacier dynamics (Jóhannesson, 1997; Jóhannes- son and others, 1995, 2006a,b,c, 2007; Aðalgeirsdót- tir, 2003; Aðalgeirsdóttir et al., 2003, 2005, 2006a, 2006b; Guðmundsson et al., 2003a,b; Marshall et al., 2005; Flowers et al., 2003, 2005). Other recent work has yielded models for describing the distribution of precipitation in Iceland (Crochet, 2007; Crochet et al., 2007; Rögnvaldsson et al., 2004, 2007; Rögnvaldsson and Ólafsson, 2005). Furthermore, glacier mass balance has been described through a degree-day model building on the following factors: temperature and precipitation outside of the glaciers, a constant temperature lapse rate, degree-day scaling factors for Figure 14. Scenario for temperature and precipitation snow and ice, and horizontal and vertical precipita- changes during the 21st century, averaged over Ice- tion gradients. land (see Jóhannesson et al., 2007). – Sviðsmynd um líklegar breytingar í hitastigi og úrkomu á 21. öld.

380 JÖKULL No. 58, 2008 Icelandic Glaciers

Figure 15. Simulated responses of Langjökull (L), Hofsjökull (H) and southern Vatnajökull (V) from 2000 through 2200 to the climate-change scenario shown in Figure 14. Volume and area reductions are normalised to present-day values, with speciﬁc runoff referring to today’s ice-covered areas and combining meltwater and precipitation. – Spár um viðbrögð jökla við loftslagsbreytingum sem sýndar eru í 14. mynd.

Plausible predictions of regional temperature and (Figure 14). Model run results for Hofsjökull, Lang- precipitation trends in Iceland have been developed jökull and southern Vatnajökull are shown in Figure in the Nordic project Climate and Energy (Rum- 15. The resulting retreat rate is similar for Hofsjök- mukainen, 2006; Bergström et al., 2007; Fenger, ull and Vatnajökull, which are predicted to lose 25% 2007; Jóhannesson et al., 2007), based on downscal- of their present volume within half a century, mean- ing of global coupled atmosphere-ocean simulations. ing that it is only on their highest peaks where ice In comparison with 1961–1990, the project scenario will survive throughout the next 200 years. Langjök- predicts a warming of 2.8◦C and a 6% increase in ull is predicted to diminish by 35% in volume over precipitation by 2071–2100. The increases in tem- 50 years and to disappear after 150 years. Consider- perature and precipitation vary according to season ing how fast Icelandic glaciers are predicted to melt in

JÖKULL No. 58, 2008 381 H. Björnsson and F. Pálsson the near future, it is not surprising that icecaps disap- Acknowledgement peared from the island during the Climatic Optimum The authors are indebted to Leó Kristjánsson, Tómas of the early Holocene. Meltwater runoff is expected Jóhannesson and Philip Vogler for improvements on to increase initially, but to peak after 40–50 years and the manuscript. then to decline to present-day values 100 years from now. The runoff increase will be highest for the lowest parts of Langjökull (∼2.8 m a−1) and next high- ÁGRIP est for Vatnajökull in parts extending nearly to sea Um 11% af Íslandi er þakið jöklum. Þeir geyma um level. The seasonal rhythm of discharge is also ex- 3 3,600 km af vatni sem janfgildir 35 m þykku lagi pected to change, with some rivers drying up entirely, jafndreifðu yfir allt landið. Þetta er stærsta vatnsforða- while others will be left to discharge exclusively pre- búr landsins og jafngildir úrkomu sem á það fellur í 20 cipitation after their glaciers have melted away. ár. Bráðni allur þessi ís hækkaði um 1 cm í heimshöf- um. Jöklarnir bregðast fljótt við sveiflum í loftslagi og hafa mikil áhrif á umhverfi sitt. Undir þeim eru mörg virk eldfjöll og frá þeim falla jökulhlaup sem ógna CONCLUSION byggð allt til sjávar. Jöklarnir veita vatni í stærstu ár landsins og frá þeim kemur nú um þriðjungur vatns Scientifically speaking, the following fields of glacio- sem rennur frá landinu. Eftir að jöklar tóku að hopa logical studies are probably the most significant re- í lok 19. aldar eftir nokkurra alda vaxtarskeið hefur garding Iceland: a) the hydrology of temperate ice- stærsti jökull landsins, Vatnajökull, minnkað um 10% 3 caps, b) interactions between glacial and volcanic að rúmmáli (300 km ) og lagt 1 mm til hækkunar sjáv- phenomena, c) glacier hazards due to jökulhlaups, d) arborðs. Undanfarin 10 ár hefur hert á jökulrýrnuninni 3 surges and the stability of ice masses, and e) the future og hann misst 2,7% af rúmmáli sínu (84 km ). Að evolution of glaciers and their role as indicators of cli- meðaltali fá jöklarnir um tvo þriðju af orku til leysing- mate change, based on the location of the island in the ar frá sól- og himingeislun og einn þriðja frá hlýju og North Atlantic Ocean, just under the Arctic Circle. röku lofti sem berst inn yfir jökulinn. Á síðasta ára- The study of all these fields is supported by unusu- tug 20. aldar bræddu eldgos og stöðugur jarðhiti um 3 ally detailed data, easily accessible on maps of glacier 5,5 km sem bættist við yfirborðsbráðnun; þar af um 3 surface and bedrock topography (using records de- 4km við Gjálpargosið. Í öllum helstu hveljöklum rived from radio-echo soundings, GPS measurements landsins berst umtalsverður hluti íss fram við fram- and satellite observations), along with a wide range of hlaup og þau hafa mikil áhrif á flæði íss og vatns frá glacier mass balance and glacio-meteorological and mörgum skriðjöklum. Sé litið á alla 20. öldina barst hydrological observations. The whole of this basic 10% af heildarísmagni til leysingarsvæða Vatnajökuls information has been applied towards increasing the með framhlaupum. Líkanreikningar af afkomu og ís- overall understanding of Icelandic glaciers and devel- flæði benda til þess að við líklegar loftslagsbreytingar oping and revising numerical models to simulate the á komandi árum muni fjórðungur til þriðjungur af nú- growth and decay of present and former glaciers and verandi rúmmáli meginjöklanna hverfa innan hálfrar to simulate the impact of climate change on glacial aldar og eftir 150 til 200 ár verði eingöngu smájöklar runoff. The current Icelandic icecaps are important eftir á hæstu fjallstindum. Afrennsli frá jöklum næði analogues for warm-based Pleistocene ice sheets and hámarki eftir um hálfa öld en yrði eftir um 100 ár jafnt may otherwise provide a suitable natural laboratory því sem nú er. Síðan minnkaði það hratt. Þegar jökl- for a variety of glaciological research. The knowledge arnir hverfa mun afrennsli til ánna berast beint frá úr- gained should have substantial international potential komu. for forecasting and comprehending the future condi- Rannsóknir á jöklum á Íslandi hafa lagt mikil- tions of today’s cold-based polar icecaps. vægan skerf til alþjóðlegra jöklarannsókna, einkum til

382 JÖKULL No. 58, 2008 Icelandic Glaciers

skilnings á vatnsrennsli um þíðjökla, framhlaupum og Rogozova, O. Sigurðsson, M. Suomalainen, Th. Thorsteins- jökulhlaupum, samspili jökla og eldvirkni, sögu lofts- son, B. Vehviläinen and N. Veijalainen 2007. Impacts of cli- lagsbreytinga og við mat á framtíð jökla á jörðinni við mate change on river runoff, glaciers and hydropower in the Nordic area. Joint final report from the CE Hydrologi- breytt veðurfar. cal Models and Snow and Ice Groups. CE Rep. No. 6, The CE Project, Reykjavík. Bergþórsson, P. 1969. An estimate of drift ice and temperature in REFERENCES Iceland in 1000 years. Jökull 19, 94–101. Aðalgeirsdóttir, G. 2003. Flow dynamics of the Vatnajökull ice Berthier, E., H. Björnsson, F. Pálsson, K. L. Feigl, M. Lubes and cap, Iceland. VAW/ETH Zürich, Mitteilungen No. 181. F. Rémy 2006. The level of the Grímsvötn subglacial lake, Vatnajökull, Iceland, monitored with SPOT5 images. Earth Aðalgeirsdóttir, G., G. H. Gudmundsson and H. Björnsson 2003. Planet. Sci. Lett. 243, 293–302. A regression model for the mass-balance distribution of the Vatnajökull Ice Cap, Iceland. Ann. Glaciol. 37, 189–193. Björnsson, H. 1971. Bægisárjökull, North-Iceland. Results of glaciological investigations 1967–1968. Part I. Mass balance Aðalgeirsdóttir, G., G. H. Guðmundsson and H. Björnsson 2005. and general meteorology. Jökull 21, 1–23. The volume sensitivity of Vatnajökull ice cap, Iceland, to per- Björnsson, H. 1972. Bægisárjökull, North-Iceland. Results of turbations in equilibrium line altitude. J. Geophys. Res. 110, glaciological investigations 1967–1968. Part II. The energy F04001, doi:10.1029/2005JF000289. balance. Jökull 22, 44–61. Aðalgeirsdóttir, G., H. Björnsson, F. Pálsson and E. Magnússon Björnsson, H. 1974. Explanation of jökulhlaups from Grímsvötn, 2006a. Analyses of a surging outlet glacier of Vatnajökull ice Vatnajökull, Iceland. Jökull 24, 1–26. cap, Iceland. Ann. Glaciol. 42, 23–28. Björnsson, H. 1975. Subglacial water reservoirs, jökulhlaups and Aðalgeirsdóttir, G., T. Jóhannesson, H. Björnsson, F. Pálsson and volcanic eruptions. Jökull 25, 1–14. O. Sigurðsson 2006b. The response of Hofsjökull and south- Björnsson, H. 1976. Marginal and supraglacial lakes in Iceland. ern Vatnajökull, Iceland, to climate change. J. Geophys. Res. Jökull 26, 40–51. 111, F03001, doi:10.1029/2005JF000388. Björnsson, H. 1978. The surface area of glaciers in Iceland. Jökull Ahlmann, H. W. 1937. Vatnajökull in relation to other present-day 28, 31. Iceland glaciers. Geografiska Annaler 19, 212–229. Björnsson, H. 1979. Glaciers in Iceland. Jökull 29, 74–80. Ahlmann, H. W. 1939. The regime of Hoffellsjökull. Geografiska Björnsson, H. 1982. Drainage basins on Vatnajökull mapped by Annaler 21, 171–188. radio echo soundings. Nordic Hydrology 1982, 213–232. Ahlmann, H. W. 1940. The relative influence of precipitation and Björnsson, H. 1983. A natural calorimeter at Grímsvötn: indica- temperature on glacier regime. Geografiska Annaler 22, 188– tions of volcanic and geothermal activity. Jökull 33, 13–18. 205. Björnsson, H. 1986a. Surface and bedrock topography of ice caps Ahlmann, H. W., and S. Þórarinsson 1937a. Object, resources and in Iceland mapped by radio echo soundings. Ann. Glaciol. 8, general progress of the Swedish-Icelandic investigations. Ge- 11–18. ografiska Annaler 19, 146–160. Björnsson, H. 1986b. Delineation of glacier drainage basins on western Vatnajökull. Ann. Glaciol. 8, 19–21. 46. Ahlmann, H. W. 1937b. Previous investigations of Vatnajökull, Björnsson, H. 1988. Hydrology of Ice Caps in Volcanic Regions. marginal oscillations of its outlet-glaciers and general de- Soc. Sci. Isl., Reykjavík, Rit 45, 139 pp. scription of its morphology. Geografiska Annaler 19, 176– Björnsson, H. 1992. Jökulhlaups in Iceland: characteristics, pre- 211. diction and simulation. Ann. Glaciol. 16, 95–106. Ahlmann, H. W. 1938. Vatnajökull. Scientific results of the Björnsson, H. 1996. Scales and rates of glacial sediment re- Swedish-Icelandic investigations 1936-37-38. The ablation. moval: A 20 km long, 300 m deep trench created beneath Geografiska Annaler 20, 171–233. Breiðamerkurjökull during the Little Ice Age. Ann. Glaciol. Ahlmann, H. W. 1939. Vatnajökull. Scientific results of the 22, 141–146. Swedish-Icelandic investigations 1936-37-38. The accumu- Björnsson, H. 1998. Hydrological characteristics of the drainage lation. Geografiska Annaler 21, 39–66. system beneath a surging glacier. Nature 395, 771–774. Bárðarson, G. G. 1934. Islands Gletscher. Beiträge zur Kenntniss Björnsson, H. 2002. Subglacial lakes and jökulhlaups in Iceland. der Gletscherbewegungen und Schwankungen auf Grund al- Global and Planetary Change 35, 255–271. ter Quellenschriften und neuesten Forschung. Rit 16, Soc. Björnsson, H., S. Björnsson and Þ. Sigurgeirsson, 1982. Sci. Isl., 60 pp. Grímsvötn: Penetration of water into hot rock boundaries of Bergström, S., T. Jóhannesson, G. Aðalgeirsdóttir, A. Ahlstrøm, magma. Nature 259, 580–581. L. M. Andreassen, J. Andréasson, S. Beldring, H. Björnsson, Björnsson, H. and H. Kristmannsdóttir 1984. The Grímsvötn B. Carlsson, P. Crochet, M. de Woul, B. Einarsson, H. Elve- geothermal area, Vatnajökull, Iceland. Jökull 34, 25–50. høy, G. E. Flowers, P. Graham, G. O. Gröndal, S. Guðmunds- Björnsson, H. and P. Einarsson 1990. Volcanoes beneath Vatna- son, S-S. Hellström, R. Hock, P. Holmlund, J. F. Jónsdóttir, jökull, Iceland: Evidence from radio echo-sounding, earth- F. Pálsson, V. Radic, N. Reeh, L. A. Roald, J. Rosberg, S. quakes and jökulhlaups. Jökull 40, 147–168.

JÖKULL No. 58, 2008 383 H. Björnsson and F. Pálsson

Björnsson, H. and M. T. Guðmundsson 1993. Variations in the Eyþórsson, J. 1931. On the present position of the glaciers in Ice- thermal output of the subglacial Grímsvötn caldera, Iceland. land. Some preliminary studies and investigations in the sum- Geophys. Res. Lett. 20, 2127–2130. mer 1930. Rit 10, Soc. Sci. Isl., Reykjavík, 35 pages. Björnsson, H., F. Pálsson, M. T. Guðmundsson and H. Haralds- Eyþórsson, J. 1935. On the variations of glaciers in Iceland. Some son 1998. Mass balance of western and northern Vatnajökull, studies made in 1931. Geografiska Annaler 17, 121–137. Iceland, 1991–1995. Jökull 45, 35–58. Eyþórsson, J. 1963. Variation of Iceland glaciers 1931–1960. Björnsson H., F. Pálsson, M. T. Guðmundsson 2000. Surface and Jökull 13, 31–33. bedrock topography of the Mýrdalsjökull ice cap, Iceland: Eyþórsson, J. 1962–1966: Jöklabreytingar (Glacier variations in The Katla caldera, eruption sites and routes of jökulhlaups. metres). Jökull 12: 37–39, 13: 29–30, 14: 97–99, 15: 148– Jökull 49, 29–46. 150. Björnsson, H., H. Rott, S. Gudmundsson, A. Fischer, A. Siegel Fenger, J., ed. 2007. Impacts of Climate Change on Renewable and M. T. Gudmundsson 2001a. Glacier-volcano interactions Energy Sources. Their Role in the Nordic Energy System. deduced by SAR interferometry. J. Glaciol. 47 (156), 58–70. A comprehensive report resulting from a Nordic Energy Re- Björnsson, H., F. Pálsson and S. Guðmundsson 2001b. Jökulsár- search project. Nord 2007:003, Nordic Council of Ministers, lón at Breiðamerkursandur, Vatnajökull, Iceland: 20th cen- Copenhagen, 190 pp. tury changes and future outlook. Jökull 50, 1–18. Fischer, A., H. Rott and H. Björnsson 2003. Observation of re- Björnsson, H., F. Pálsson and H. Haraldsson 2002. Mass balance cent surges of Vatnajökull, Iceland, by means of ERS SAR of Vatnajökull (1991–2001) and Langjökull (1996–2001), interferometry. Ann. Glaciol. 37, 69–76. Iceland. Jökull 51, 75–78. Flowers, G. E., H. Björnsson and F. Pálsson 2003. New insights Björnsson, H., F. Pálsson, O. Sigurðsson and G. E. Flowers 2003. into the subglacial and periglacial hydrology of Vatnajök- Surges of glaciers in Iceland. Ann. Glaciol. 36, 82–90. ull, Iceland, from a distributed physical model. J. Glaciol. Björnsson, H., S. Guðmundsson and F. Pálsson 2005. Glacier 49 (165), 257–270. winds on Vatnajökull ice cap, Iceland and their relation to Flowers, G. E., H. Björnsson, F. Pálsson and G. K. temperatures of its lowland environs. Ann. Glaciol. 42, 291– C. Clarke 2004. A coupled sheet-conduit model of 296. jökulhlaup propagation. Geophys. Res. Lett. 31, L05401, Black, J., G. Miller, Á. Geirsdóttir, W. Manley and H. Björnsson doi:10.1029/2003GL019088. 2004. Sediment thickness and Holocene erosion rates from a Flowers, G. E., S. J. Marshall, H. Björnsson and G. K. C. Clarke seismic survey of Hvítárvatn, central Iceland. Jökull 54, 37– 2005. Sensitivity of Vatnajökull ice cap hydrology and dy- 56. namics to climate warming over the next two centuries. J. Caseldine, C. J. and J. Stötter 1993. “Little Ice Age” glaciation Geophys. Res. 110, F02011, doi:10.1029/2004JF000200. of Trollaskagi peninsula, northern Iceland: climatic implica- tions for reconstructed equilibrium line altitudes (ELAs). The Flowers, G. E., H. Björnsson, Á. Geirsdóttir, G. H. Miller and G. Holocene 3,357–366. K. C. Clarke 2007. Glacier fluctuation and inferred climatol- ogy of Langjökull ice cap through the Little Ice Age. Quater. Crochet, P. 2007. A study of regional precipitation trends in Ice- Sci. Rev. 22, 2337–2353. land using a high quality gauge network and ERA-40. J. Cli- mate 20 (18), 4659–4677, doi:10.1175/JCLI4255.1. Flowers, G. E., H. Björnsson, Á. Geirsdóttir, G. H. Miller, J. L. Crochet, P., T. Jóhannesson, T. Jónsson, O. Sigurðsson, H. Björns- Black and G. K. C. Clarke 2008. Holocene climate conditions son, F. Pálsson and I. Barstad 2007. Estimating the spatial and glacier variations in central Iceland from physical mod- distribution of precipitation in Iceland using a linear model eling and empirical evidence. Quater. Sci. Rev. 27, 797–813. of orographic precipitation. J. Hydrometeorol. 8 (6), 1285– Gudmundsson, G. H., G. Aðalgeirsdóttir and H. Björnsson 2003. 1306. Observational verification of predicted increase in bedrock- Dugmore, A. J. 1989. Tephrochronological studies of Holocene to-surface amplitude transfer during a glacier surge. Ann. glacier fluctuations in south Iceland. In: J. Oerlemans, ed., Glaciol. 36, 91–96. Glacier Fluctuations and Climatic Change, Kluwer Aca- Gudmundsson, H. J. 1997. A review of the Holocene environmen- demic Publishers, Dordrecht, Netherlands, 37–55. tal history of Iceland. Quater. Sci. Rev. 16(1), 81–92. Einarsson, M. Á. 1984. Climate of Iceland. In: van Loon, H., ed., Guðmundsson, M. T., H. Björnsson and F. Pálsson 1995. Changes Climates of the Oceans, Elsevier, Amsterdam, 673–697. in jökulhlaup sizes in Grímsvötn, Vatnajökull, Iceland, 1934– Einarsson, Þ. 1963. Pollen analytical studies on the vegetation and 1991, deduced from in situ measurements of subglacial lake climate history of Iceland in Late and Postglacial times. In: volume. J. Glaciol. 41 (138), 263–272. A. Löve and D. Löve, eds. North Atlantic Biota and Their Guðmundsson, M. T., F. Sigmundsson and H. Björnsson 1997. History, Pergamon Press, Oxford, 355–365. Ice-volcano interaction of the 1996 Gjálp subglacial eruption, Eiríksson, J., K. L. Knudsen, H. Hafliðason and J. Heinemeier Vatnajökull, Iceland. Nature 389, 954–957. 2000. Chronology of late Holocene climatic events in the Guðmundsson, M. T., F. Sigmundsson, H. Björnsson and Th. northern North Atlantic based on AMS 14C dates and tephra Högnadóttir 2003. The 1996 eruption at Gjálp, Vatnajökull markers from the volcano Hekla, Iceland, J. Quater. Sci. 15, ice cap, Iceland: efficiency of heat transfer, ice deformation 573–580. and subglacial water pressure. Bull. Volc. 66, 46–65.

384 JÖKULL No. 58, 2008 Icelandic Glaciers

Guðmundsson, S., M. T. Gudmundsson, H. Björnsson, F. Sig- Kirkbride, M. P. 2002. Icelandic climate and glacier fluctuations mundsson, H. Rott and J. M. Carstensen 2002. Three- through the termination of the “Little Ice Age”, Polar Geog- dimensional glacier surface motion maps at the Gjálp erup- raphy 26,116–133. tion site, Iceland, inferred from combining InSAR and other Kirkbride, M. P. and A. J. Dugmore 2001. Timing and signifi- ice displacement data. Ann. Glaciol. 34, 315–322. cance of mid-Holocene glacier advances in northern and cen- Guðmundsson, S., H. Björnsson, F. Pálsson and H. H. Haralds- tral Iceland, J. Quater. Sci. 16,145–153. son 2006. Energy balance of Brúarjökull and circumstances Kirkbride, M. P. and A. J. Dugmore 2006. Responses of moun- leading to the August 2004 floods in the river Jökla, N- tain ice caps in central Iceland to Holocene climate change. Vatnajökull. Jökull 55, 121–138. Quater. Sci. Rev. 25 (13–14), 1692–1707. Hanna, E., T. Jónsson and J. E. Box 2004. An analysis of Ice- Magnússon, E., H. Björnsson, F. Pálsson and J. Dall 2004. landic climate since the nineteenth century. Int. J. Climatol. Glaciological application of InSAR topography data of W- 24, 1193–1210. Vatnajökull acquired in 1998. Jökull 54, 17–36. Jóhannesson, T. 1986. The response time of glaciers in Iceland to Magnússon, E., H. Björnsson and F. Pálsson 2005a. Íssjármælin- changes in climate, Ann. Glaciol. 8, 100–101. gar á Drangajökli 8.–9. apríl 2004. (Radio-echo soundings on Jóhannesson, T. 1997. The response of two Icelandic glaciers Drangajökull, 2004., in Icelandic), Jökull 54, 85–86. to climatic warming computed with a degree-day glacier Magnússon, E., H. Björnsson, J. Dall and F. Pálsson 2005b. Vol- mass balance model coupled to a dynamic glacier model. J. ume changes of Vatnajökull ice cap, Iceland, due to surface Glaciol. 43 (143), 321–327. mass balance, ice flow, and sub-glacial melting at geothermal areas. Geophys. Res. Lett. 32 (5), L05504. Jóhannesson, T. 2002. Propagation of a subglacial flood wave dur- Magnússon, E., H. Björnsson, J. Dall and F. Pálsson 2005c. ing the initiation of a jökulhlaup. Hydrol. Sci. J. 47 (3), 417– The 20th century retreat of ice caps in Iceland derived from 434. airborne SAR: W-Vatnajökull and N-Mýrdalsjökull. Earth Jóhannesson, T., O. Sigurðsson, T. Laumann and M. Kennett Planet. Sci. Lett. 237, 508–515. 1995. Degree-day glacier mass balance modelling with ap- Magnússon E., H. Björnsson and F. Pálsson 2007. Radio plications to glaciers in Iceland, Norway and Greenland. J. echo sounding on Kvíárjökull, Hrútárjökull and Fjallsjökull. Glaciol. 41 (138), 345–358. Jökull 54, 85–86. Jóhannesson, T. and O. Sigurðsson 1998. Interpretation of glacier Marshall, S. J., H. Björnsson, G. E. Flowers and G. K. C. Clarke variations in Iceland 1930–1995. Jökull 45, 27–33. 2005. Simulation of Vatnajökull ice cap dynamics. J. Geo- Jóhannesson, T., O. Sigurðsson, B. Einarsson and Th. Thorsteins- phys. Res. 110, F03009, doi:10.1029/2004JF000262. son 2006a. Mass balance modeling of the Hofsjökull ice cap Nick, F. M., J. van der Kwast and J. Oerlemans 2007. Simulation based on data from 1988–2004. Rep. OS-2006/004, Nat. En- of the evolution of Breiðamerkurjökull in the late Holocene. ergy Auth., Reykjavík. J. Geophys. Res. 112, B01103, doi:10.1029/2006JB004358. Jóhannesson, T., G. Aðalgeirsdóttir, A. Ahlstrøm, L. M. An- Ogilvie, A. E. J. 1992. Documentary evidence for changes in the dreassen, H. Björnsson, M. de Woul, H. Elvehøy, G. E. Flow- climate of Iceland, A.D. 1500 to 1800. In: R. S. Bradley and ers, S. Guðmundsson, R. Hock, P. Holmlund, F. Pálsson, V. P. D. Jones, eds., Climate since A.D. 1500, Routledge, Lon- Radic, O. Sigurðsson and Th. Thorsteinsson 2006b. The im- don and New York, 92–117. pact of climate change on glaciers and glacial runoff in the Ogilvie, A. E. J. and T. Jónsson 2001. “Little ice age” research: Nordic countries. Proc. European Conference of Impacts of A perspective from Iceland. Climatic Change 48 (1), 9–52, Climate Change on Renewable Energy Sources, Reykjavík, doi:10.1023/A:1005625729889. Iceland. Oerleman, J., H. Björnsson, M. Kuhn, F. Obleitner, F. Páls- Jóhannesson, T., G. Aðalgeirsdóttir, H. Björnsson, C. E. Bøggild, son, P. Smeets, H. F. Vugts and J. de Wolde 1999. A H. Elvehøy, S. Guðmundsson, R. Jóhannesson, H. Björnsson, glacio-meteorological experiment on Vatnajökull, Iceland. P. Crochet, F. Pálsson, O. Sigurðsson and Th. Thorsteinsson Boundary-Layer Meteorology 92, 3–26. 2006c. Mass balance modeling of the Vatnajökull, Hofsjök- Pálsson, S. and G. H. Vigfússon 1996. Gagnasafn aurburðar- ull and Langjökull ice caps. Proc. European Conference of mælinga 1963–1995. Orkustofnun OS-96032/VOD-05, B Impacts of Climate Change on Renewable Energy Sources, 270 pp. Reykjavík, Iceland. Pálsson, S., J. Harðardóttir, G. H. Vigfússon and Á. Snorra- Jóhannesson, T., G. Aðalgeirsdóttir, H. Björnsson, P. Crochet, E. son 2000. Reassessment of suspended sediment load of river B. Elíasson, S. Guðmundsson, J. F. Jónsdóttir, H. Ólafsson, F. Jökulsá á Dal at Hjardarhagi. Orkustofnun, Vatnamælingar Pálsson, Ó. Rögnvaldsson, O. Sigurðsson, A. Snorrason, Ó. OS-2000/070, 30 pp. G. Bl. Sveinsson and Th. Thorsteinsson 2007. Effect of cli- Rist, S. 1955. Skeiðarárhlaup 1954. Jökull 15, 30–36. mate change on hydrology and hydro-resources in Iceland. Rist, S. 1956. Íslenzk vötn, (Icelandic Fresh Waters). Raforku- Final report of the VO-project, OS-2007/011. málastjóri. Vatnamælingar. Reykjavík, 127 pp. Jónsdóttir, J. F. 2008. A runoff map based on numerically sim- Rist, S. 1970. Annáll um jökulhlaup. Jökull 20, 89–90. ulated precipitation and a projection of future runoff in Rist, S. 1973. Jökulhlaupaannáll 1971, 1972 og 1973. Jökull 23, Iceland. Hydrol. Sci. J. 53(1), 100–111, doi:10.1623/hysj.- 55–60. 53.1.100. Rist, S. 1976. Grímsvatnahlaupið 1976. Jökull 26, 80–90.

JÖKULL No. 58, 2008 385 H. Björnsson and F. Pálsson

Rist, S. 1981. Jökulhlaupaannáll 1977, 1978, 1979 og 1980. Jökull Stötter, J., M. Wastl, C. Caseldine and T. Häberle 1999. 30, 31–35. Holocene palaeoclimatic reconstructions in northern Iceland: Rist, S. 1984. Jökulhlaupaannáll 1981, 1982 og 1983. Jökull 34, approaches and results, Quater. Sci. Rev. 18, 457–474. 165–179. Sverrisson, M., Æ. Jóhannesson and H. Björnsson 1980. Radio- Rist, S. 1967–1987. Jöklabreytingar (Glacier variations in me- echo equipment for depth sounding of temperate glaciers. J. tres). Jökull 17: 321–325, 18: 401–405, 20: 83–87, 21: 73– Glaciol. 93, 477–485. 77, 22: 89–95, 23: 61–66, 24: 77–82, 25: 73–79, 26: 69–74, Tómasson, H. 1973. Hamfarahlaup í Jökulsá á Fjöllum. Náttúru- 27: 88–93, 28: 61–65, 31: 37–46, 32: 121–125, 33: 141–144, fræðingurinn 43, 12–34. 34: 173–179, 35: 111–119, 36: 83–90, 37: 85–90. Tómasson, H. 1974. Grímsvatnahlaup 1972. Mechanism and sed- Rummukainen, M. 2006. The CE regional climate scenarios. iment discharge. Jökull 24, 27–39. Proc. European Conference of Impacts of Climate Change Tómasson, H. 1981. Vatnsafl Íslands, mat á stærð orkulindar (Hy- on Renewable Energy Sources, Reykjavík, Iceland, June 5–6. dropower in Iceland, an estimate of the size of an energy Rögnvaldsson, Ó., P. Crochet and H. Ólafsson 2004. Mapping source. In: Orkuþing 81. Vol. 2. Reykjavík. of precipitation in Iceland using numerical simulations and Tómasson, H. 1982. Vattenskraft i Island och dess hydrologiska statistical modeling. Meteorol. Zeitschrift 13 (3), 209–219, förutsättingar (Hydropower in Iceland and its hydrological doi:10.1127/0941-2948/2004/0013-0209. prerequisites). In: Den nordiske hydrologiske konference, Rögnvaldsson, Ó. and H. Ólafsson 2005. The response of precipi- NHK-92, Förde, Orkustofnun, OS-82059/VOD-10. tation to orography in simulations of future climate. Proc. Int. Conf. Alpine Meteorol. May 2005, Zadar, Croatia (ICAM), Tómasson, H. 1996. The jökulhlaup from Katla in 1918. Ann. Croatian Meteorol. Journal 40, 526–529. Glaciol. 22, 249–254. Rögnvaldsson, Ó., J. F. Jónsdóttir and H. Ólafsson 2007. Nu- Tómasson, H., S. Pálsson and P. Ingólfsson 1980. Comparison of merical simulations of precipitation in the complex terrain sediment load in the Skeiðará jökulhlaups in 1972 and 1976. of Iceland-Comparison with glaciological and hydrological Jökull 30, 21–33. data. Meteorol. Zeitschrift 16 (1), 71–85. Vinther, B. N., H. B. Clausen, S. J. Johnsen, S. O. Rasmussen, K. Schomacker, A., J. Krüger and G. Larsen 2003. An extensive Late K. Andersen, S. L. Buchardt, I. K. Seierstad, M. L. Siggaard- Holocene glacier advance of Kötlujökull, central south Ice- Andersen, J. P. Steffensen, A. M. Olsen and J. Heinemeier land. Quater. Sci. Rev. 22, 1427–1434. 2006. A synchronized dating of three Greenland ice cores Sharp, M. J. and A. J. Dugmore 1985. Holocene glacier fluctu- throughout the Holocene. J. Geophys. Res. 111, D13102. ations in eastern Iceland. Zeitschrift für Gletscherkunde und Wastl, M., J. Stötter and C. Caseldine 2001. Reconstruction Glazialgeologie 21, 341–349. of Holocene variations of the upper limit of tree or shrub Sigurðsson, O. 1998. Glacier variations in Iceland 1930–1995 – Birch growth in northern Iceland based on evidence from from the database of the Iceland Glaciological Society. Jökull Vesturárdalur-Skiðadalur, Tröllaskagi. Arctic, Antarctic and 45, 3–25. Alpine Res. 33, 191–203. Sigurðsson, O. and Ó. J. Sigurðsson 1998. Afkoma nokkurra jökla Þórarinsson, S. 1939. The ice dammed lakes of Iceland with par- á Íslandi 1992–1997. (Mass balance of a number of Icelandic ticular reference to their value as indicators of glacier oscil- glaciers 1992–1997). Rep. OS-98082 (in Icelandic), Nat. En- lations. Geografiska Annaler 21, 216–242. ergy Auth., Reykjavík. Þórarinsson, S. 1943. Oscillations of the Iceland glaciers during Sigurðsson, O. 2005. Variations of termini of glaciers in Iceland in the last 250 years. Geografiska Annaler 25, 1–54. recent centuries and their connection with climate. In: Casel- Þórarinsson, S. 1964. On the age of the terminal moraines of dine, C., A. Russell, J. Harðardóttir and Ó. Knudsen, eds., Brúarjökull and Hálsajökull, Jökull 14, 67–75. Iceland – Modern Processes and Past Environments. Else- Þórarinsson, S. 1966. The age of the maximum postglacial ad- vier, Amsterdam, 241–255. vance of Hagafellsjökull Eystri. Jökull 16, 207–210. Sigurðsson, O., Th. Thorsteinsson, S. M. Ágústsson and B. Ein- Þórarinsson, S. 1969. Glacier surges in Iceland with special refer- arsson 2004. Afkoma Hofsjökuls 1997–2004. (Mass balance ence to surges of Brúarjökull. Can. J. Earth Sci. 6, 875–882. of Hofsjökull 1997–2004). Rep. OS-2004/029, Nat. Energy Þórarinsson, S. 1974. Vötnin stríð. Saga Skeiðarárhlaupa og Auth., Reykjavík. Grímsvatnagosa, (Jökulhlaups in Skeiðará and Grímsvötn Sigurðsson, O., T. Jónsson and T. Jóhannesson 2007. Relation be- eruptions, in Icelandic). Bókaútgáfa Menningarsjóðs, tween glacier-termini variations and summer temperature in Reykjavík. 254 pages. Iceland since 1930. Ann. Glaciol. 42, 395–401. Stötter, J. 1991. New observations on the postglacial glacial his- Þórarinsson, S. 1975. Katla og annáll Kötlugosa. (Katla eruptions, tory of Tröllaskagi, Northern Iceland. In: J. K. Maizels and C. in Icelandic), Árbók Ferðafélags Íslands, Reykjavík, 125– Caseldine, eds., Environmental Change in Iceland: Past and 149. Present, Kluwer Academic Publishers, Dordrecht, Nether- lands, 181–192.

386 JÖKULL No. 58, 2008