Papers by Ólafur Ingólfsson
Boreas, 2008
... btgdfsson urrd I . iw &rrnekow. D(prrrrtic,rit o / Qutrrert~tir~ ... more ... btgdfsson urrd I . iw &rrnekow. D(prrrrtic,rit o / Qutrrert~tir~ G'rolu~~~. Lurid Ur~iiw.~iIj. Tornur. 13, S-?L?? 63 I-iind. Sbre(/rtr: o/fJ Brrrnikr. G'i~olo~ii~ul Surri?j of' Denrtiurk. Tl~oruwj 8. DK-24400 Coipc~nhugcw N V. Dcwmurk; Do~~icl N. Peritr),. C;eologi.sk lti.\~i/rir. ...
Boreas, 2009
The morphology, sedimentology and architecture of an end moraine formed by a ∼9 km surge of Brúar... more The morphology, sedimentology and architecture of an end moraine formed by a ∼9 km surge of Brúarjökull in 1963–64 are described and related to ice-marginal conditions at surge termination. Field observations and accurate mapping using digital elevation models and high-resolution aerial photographs recorded at surge termination and after the surge show that commonly the surge end moraine was positioned underneath the glacier snout by the termination of the surge. Ground-penetrating radar profiles and sedimentological data reveal 4–5 m thick deformed sediments consisting of a top layer of till overlying gravel and fine-grained sediments, and structural geological investigations show that the end moraine is dominated by thrust sheets. A sequential model explaining the formation of submarginal end moraines is proposed. The hydraulic conductivity of the bed had a major influence on the subglacial drainage efficiency and associated porewater pressure at the end of the surge, thereby affecting the rates of subglacial deformation. High porewater pressure in the till decreased its shear strength and raised its strain rate, while low porewater pressure in the underlying gravel had the opposite effect, such that the gravel deformed more slowly than the till. The principal velocity component was therefore located within the till, allowing the glacier to override the gravel thrust sheets that constitute the end moraine. The model suggests that the processes responsible for the formation of submarginal end moraines are different from those operating during the formation of proglacial end moraines.
The internal architecture and structural evolution of the Arnarfellsmúlar terminal moraine at Múl... more The internal architecture and structural evolution of the Arnarfellsmúlar terminal moraine at Múlajökull, a surge-type glacier in central Iceland, is described in order to demonstrate submarginal and proglacial glaciotectonic processes during glacier surging, as well as constraining the age of the maximum extent of the glacier. The moraine is 4–7 m high, 50–100 m wide, and composed of a highly deformed sequence of loess, peat, and tephra that is draped by till up to the crest. The internal architecture is dominated by steep, high-amplitude overturned folds and thrusts in the crest zone but open, low-amplitude folds on the distal slope. Section balancing suggests a basal detachment (décollement) depth of 1.4 m and a total horizontal shortening of around 59%. This implies that the glacier coupled to the foreland about 70 m up glacier from its terminal position to initiate the formation of the moraine. The structural evolution is polyphase in that the formation commenced with low-amplitude open folding of the foreland, followed by overfolding and piggyback thrusting. Radiocarbon dating and analysis of tephra layers, along with historical references, indicate that the most likely time of moraine formation was between A.D. 1717 and 1760, which suggests that Múlajökull had its Little Ice Age maximum and most extensive surge earlier than many other surge-type glaciers in Iceland.
Time-series of digital elevation models (DEMs) of the forefield of the Brúarjökull surge-type gla... more Time-series of digital elevation models (DEMs) of the forefield of the Brúarjökull surge-type glacier in Iceland were used to quantify the volume of material that was mobilized by the 1963–1964 surge. The DEMs were produced by stereophotogrammetry on aerial photographs from before the surge (1961) and after (1988 and 2003). The analysis was performed on two DEMs of Difference (DoDs), i.e., a 1961–2003 DoD documenting the impact of the surge and a 1988–2003 DoD documenting the post-surge modification of the juvenile surging glacier landsystem. Combined with a digital geomorphological map, the DoDs allow us to quantify the impact of the surge on a landsystem scale down to individual landforms. A total of 34.2 ± 11.3 × 10 6 m 3 of material was mobilized in the 30.7-km 2 study area as a result of the most recent surge event. Of these, 17.4 ± 6.6 × 10 6 m 3 of the material were eroded and 16.8 ± 4.7 × 10 6 m 3 were deposited. More than half of the deposited volume was ice-cored landforms. This study demonstrates that although the total mobilized mass volume is high, the net volume gain of ice and sediment deposited as landforms in the forefield caused by the surge is low. Furthermore, depo-sition of new dead-ice from the 1963–1964 surge constitutes as much as 64% of the volume gain in the forefield. The 1988–2003 DoD is used to quantify the melt-out of this dead-ice and other paraglacial modification of the recently deglaciated forefield of Brúarjökull.
Surging glaciers are potential analogues for land-terminating palaeo-ice streams and surging ice ... more Surging glaciers are potential analogues for land-terminating palaeo-ice streams and surging ice sheet lobes, and research on surge-type glaciers is important for understanding the causal mechanisms of modern and past ice sheet instabilities. The geomorphic signatures left by the Icelandic surge-type glaciers vary and range from glaciotectonic end moraines formed by folding and thrusting, crevasse-squeeze ridges, concertina eskers, drum-lins and fluted forefields, to extensive dead-ice fields and even drift sheets where fast ice-flow indicators are largely missing. We outline some outstanding research questions and review case studies from the surge-type outlets of Brúarjökull, Eyjabakkajökull and Tungnaárjökull (Vatnajökull ice cap), Múlajökull and Sátujökull (Hofsjökull ice cap), Hagafellsjökull and Suðurjökull (Langjökull ice cap), Kaldalónsjökull, Leirufjarðarjökull and Reykjarfjarðarjökull (Drangajökull ice cap), as well as the surge-type cirque glaciers in northern Iceland. We review the current understanding of how rapid ice flow is sustained throughout the surge, the processes that control the development of the surge-type glacier landsystem and the geological evidence of surges found in sediments and landforms. We also examine if it is possible to reconstruct past surge flow rates from glacial landforms and sediments and scale-up present-day surge processes, landforms and landsystems as modern analogues to past ice streams. Finally, we also examine if there is a climate/mass-balance control on surge initiation, duration and frequency.
Quaternary Science Reviews, 2008
is furnished to the author for non-commercial research and education use, including for instructi... more is furnished to the author for non-commercial research and education use, including for instruction at the author's institution, sharing with colleagues and providing to institution administration.
The drumlin field at Múlajökull, Iceland, is considered to be an active field in that partly and ... more The drumlin field at Múlajökull, Iceland, is considered to be an active field in that partly and fully ice-covered drumlins are being shaped by the current glacier regime. We test the hypothesis that the drumlins form by a combination of erosion and deposition during successive surge cycles. We mapped and measured 143 drumlins and studied their stratigraphy in four exposures. All exposures reveal several till units where the youngest till commonly truncates older tills on the drumlin flanks and proximal slope. Drumlins inside a 1992 moraine are relatively long and narrow whereas drumlins outside the moraine are wider and shorter. A conceptual model suggests that radial crevasses create spatial heterogeneity in normal stress on the bed so that deposition is favoured beneath crevasses and erosion in adjacent areas. Consequently, the crevasse pattern of the glacier controls the location of proto-drumlins. A feedback mechanism leads to continued crevassing and increased sedimentation at the location of the proto-drumlins. The drumlin relief and elongation ratio increases as the glacier erodes the sides and drapes a new till over the landform through successive surges. Our observations of this only known active drumlin field may have implications for the formation and morphological evolution of Pleistocene drumlin fields with similar composition, and our model may be tested on modern drumlins that may become exposed upon future ice retreat.
Stratigraphic and morphologic data previously collected from the forefield of Múlajökull, Iceland... more Stratigraphic and morphologic data previously collected from the forefield of Múlajökull, Iceland, suggest that its recent surge cycles are responsible for the formation of drumlins there and that their relief reflects both deposition on drumlins and erosion between them. We have tested these ideas and aspects of leading models of drumlin formation by studying past patterns of bed deformation and effective stress in basal tills of the glacier's forefield. Patterns of till strain indicated by the anisotropy of magnetic susceptibility (AMS) of ~2300 intact till samples indicate that till was deposited during shear deformation, with shearing azimuths and planes that conform to the drumlin morphology. Thus, till deposition occurred as drumlins grew, in agreement with LiDAR data indicating that the degree of aggradation of the glacier forefield is largest in areas subjected to the most surges. Previously described unconformities on the drumlin flanks, however, indicate that drumlin relief at Múlajökull has resulted, in part, from erosion. Given that the last surge deposited a till layer both on and between drumlins, a reasonable hypothesis is that erosion between drumlins occurred during normal (quiescent) flow of the glacier between surges. Densities of till samples , analyzed in conjunction with laboratory consolidation tests, indicate that effective stresses on the bed during such periods were on the order of 100 kPa larger between drumlins than within them, an observation consistent with subglacial channels at low water pressure occupying interdrumlin areas. Transport of sediment by turbulent flow in these channels or high effective stress adjacent to them causing enhanced till entrainment in ice or increased depths of bed deformation would promote the sediment flux divergence necessary to erode areas between drumlins. The observation that effective stresses were higher between drumlins than within them is the opposite of that presumed in leading models of drumlin formation. Moreover, the lack of AMS-fabric evidence of longitudinal compression in drumlin tills does not support some models of drumlin formation that invoke negative till-flux gradients in a deforming bed.
Keywords: Kalahari Dwyka Formation Rehoboth Province Kaapvaal Craton Ion probe zircon dating Zirc... more Keywords: Kalahari Dwyka Formation Rehoboth Province Kaapvaal Craton Ion probe zircon dating Zircon oxygen isotope We have found a new source of information about what lies beneath the Kalahari sands. The Kheis and Rehoboth Provinces of southern Africa were thought to be underlain by either an ~1800 Ma orogenic belt, or a northern branch of the ~ 1200 Ma Namaqua–Natal Province. Glacial diamictites of the Permocarboniferous Dwyka Group exposed at Rietfontein west of the Kalahari sands carry cobbles plucked from the bedrock by the ice sheet which covered the Gondwana supercontinent about 300 Ma ago. Despite altered mineralogy, the cobbles are not severely altered geochemically. Their normative mineral compositions give classifications as trondhjemites and granites, supported by rare earth element diagrams. Microbeam U–Pb zircon dating of the granitic cobbles shows that they contain no evidence of crustal growth or orogeny at either 1800 or 1200 Ma. Rather they testify to the presence of 2500 to 2900 Ma Archaean trondhjemitic and granitic crust beneath the Kalahari, with a lesser ~ 2050 Ma granite component. The pebble assemblages from the diamictites we sampled lack the diagnostic banded iron formation (BIF), stromatolitic limestone and other supracrustal pebbles which characterise diamictites derived from the Kaapvaal Craton, thus we envisage shorter transport distances and derivation from the region now beneath the Kalahari sands. Three of the Archaean granite cobbles have unusual less-than-mantle zircon oxygen isotope values around + 3 (δ 18 O VSMOW), which may reflect interaction of their source with high-temperature, originally meteoric water before melting to produce the granites. The mafic cobbles described in a companion paper are much younger and are related to intrusions of the 1.1 Ga Umkondo Large Igneous Province, probably located on the Kalahari Line or Rehoboth Province. Five trondhjemitic granites from the westernmost outcrops of the Kaapvaal Craton were dated, the oldest being 3061 ± 9 Ma and four others between 2882 ± 7 Ma and 2854 ± 7 Ma, reflecting the cratonisation of the Kimberley Terrane. Four of the Archaean Dwyka cobbles we dated are younger than the 2.7 Ga Kaapvaal cover sequence and are thus too young to be derived from the craton. All the Dwyka cobbles described here are most likely derived from either the Rehoboth Province or the Kalahari Line with origins from the Kheis Province, Kaapvaal Craton, or further afield considered unlikely. We envisage the Rehoboth Province to consist of an Archaean core supplemented by Palaeoproterozoic granitoids, which was joined to the Kaapvaal Craton at an early stage of crustal development and played an important role during later tectonic events. This has important implications not only for the tectonic framework and assembly of Southern Africa, but also for exploration for diamonds and other ore deposits.
The maximum limits of the Eurasian ice sheets during four glaciations have been reconstructed: (1... more The maximum limits of the Eurasian ice sheets during four glaciations have been reconstructed: (1) the Late Saalian (>140 ka), (2) the Early Weichselian (100–80 ka), (3) the Middle Weichselian (60–50 ka) and (4) the Late Weichselian (25–15 ka). The reconstructed ice limits are based on satellite data and aerial photographs combined with geological field investigations in Russia and Siberia, and with marine seismic-and sediment core data. The Barents-Kara Ice Sheet got progressively smaller during each glaciation, whereas the dimensions of the Scandinavian Ice Sheet increased. During the last Ice Age the Barents-Kara Ice Sheet attained its maximum size as early as 90–80,000 years ago when the ice front reached far onto the continent. A regrowth of the ice sheets occurred during the early Middle Weichselian, culminating about 60–50,000 years ago. During the Late Weichselian the Barents-Kara Ice Sheet did not reach the mainland east of the Kanin Peninsula, with the exception of the NW fringe of Taimyr. A numerical ice-sheet model, forced by global sea level and solar changes, was run through the full Weichselian glacial cycle. The modeling results are roughly compatible with the geological record of ice growth, but the model underpredicts the glaciations in the
The Nordic countries have experienced multiple glacia-tions and intervening interglacials during ... more The Nordic countries have experienced multiple glacia-tions and intervening interglacials during the last ca. 2.5–3 million years. Although evidence from Greenland and Ice-land shows that ice sheets started to expand some time before 3 Ma, little is known about the glaciations and intervening interglacials older than the last Glacial Maximum due to repeated phases of glacial erosion and reworking. The extensive Saalian glaciation (c. 140 ka BP) contributed to high sea levels in Greenland and in the Baltic area during the early part of the last interglacial (Eemian). Temperatures were about 5 ºC higher during the Eemian than they are today and the Greenland ice sheet was reduced to about half of its present size, causing globally higher sea levels than we have today. Ice extent in Fennoscandia was restricted during early Weichselian sta-dials, but middle Weichselian ice advances in Scandinavia reached as far as Denmark. During the Last Glacial Maximum , large ice sheets were present in all Nordic countries and coalesced with neighboring ice sheets. Deglaciation commenced around 17–15 ka BP in most areas and was promoted by rapidly rising global sea level and glacial isostasy. The Younger Dryas cold event(c. 12.6–11.5 ka BP) is seen as a short-term re-advance, still-stand or fluctuation of land-based ice sheet margins. Around 7–9 ka BP ice sheets had disappeared or had attained their present size. While uplift is still going on in some regions, others are subject to submergence. The different stages of development of the Baltic Sea are an example of how the intricate interplay between glacial eustasy and isostasy influences sedimentation, basin size and drainage patterns.
— At the Last Glacial Maximum (LGM) glaciers extended out towards the shelf break around Iceland ... more — At the Last Glacial Maximum (LGM) glaciers extended out towards the shelf break around Iceland and ice thickness over most of Iceland was 2000 m. Rapidly rising global sea level at 13.0–12.5 ka BP caused extremely fast deglaciation and collapse of the marine based part of the Icelandic ice sheet. Raised shorelines at very high altitudes, dated to about 12.6 ka BP, signify the rapid glacial retreat and transgression of relative sea level in coastal areas. A readvance of the Icelandic ice sheet culminated in Younger Dryas times, about 10.3 ka BP, when the ice margin was close to the present coastline. Following ice retreat and regression of relative sea level at the end of the Younger Dryas, the ice sheet advanced again and reached a spatial extent close to its Younger Dryas maximum at about 9.8 ka BP. Subsequently, the ice sheet retreated rapidly and at about 9.4 ka BP relative sea level fell towards and eventually below present sea level. Relative sea level had regressed to about 40 m below present sea level at about 8.9 ka BP, when the rate of isostatic rebound was slowing down and eustatic sea level rise caused the onset of a transgression. At about 8.0 ka BP Icelandic glaciers were of a similar or a little lesser extent than at present, and during the mid-Holocene thermal optimum some of the present-day ice caps may have been signicantly reduced or absent. The onset of Neoglaciation occurred after 6–5 ka BP and most Icelandic glaciers reached their Holocene maximum during the Little Ice Age. The vegetation history of Iceland reects the glacial history in that the earliest evidence of plant succession is from mid Allerød times as grass and dwarf shrub tundra developed in the wake of the initial ice retreat. The younger Dryas cooling is evident in the biostratigraphical record as discontinuous vegetation cover and tundra environments developed. At the beginning of the Holocene, dwarf shrub, and later on, shrub heath with Salix, Dwarf birch, and Juniperus became established. Birch woodland, the climax vegetation during the Holocene, most probably had its greatest extent during the Atlantic Chronozone, at about 7–6 ka BP. The woodland showed a retrogressive succession towards more open landscape during the second half of the Holocene as widespread mires, dwarf shrub heaths and grassland took over in the landscape evolution. The Norse settlement of Iceland in the 9th Century AD caused rapid vegetation changes in the wake of the settlement, where primarily intensive grazing caused woodland destruction and expansion of grass heath, dwarf shrub heath and mires.
The deglaciation of the Late Weichselian Ice-landic ice sheet was extremely rapid, and marked by ... more The deglaciation of the Late Weichselian Ice-landic ice sheet was extremely rapid, and marked by a collapse of its marine-based sections between 15.0 and 14.7 ka BP. We present a conceptual two-dimensional glacio-isostatic equilibrium model that describes the effect of simultaneous glacio-isostatic responses to changes in glacier load on the crust and relative eustatic sea level changes during the deglaciation of the Icelandic ice sheet. The deglaciation was characterized by three main steps: (1) slow deglaciation of the outer shelf, progressing between about 18.6–15.0 ka BP, mainly driven by sea-level rise and grounding line retreat on a retrograde slope; (2) an extremely rapid retreat from the shelf areas between 15.0 and 14.7 ka BP, mainly driven by rapid sea-level rise and large-scale calving; (3) a slower retreat driven by North Atlantic warming once the ice sheet was inside the coast, between 14.7 and 13.8 ka BP. The very rapid deglaciation of the Iceland shelf areas between 15.0 and 14.7 ka BP does not confirm with linear response to warming, but rather describes a non-linear response, where the ice sheet experiences an abrupt collapse once a threshold defined by relative eustatic sea-level rise is crossed. We propose that rapid eustatic sea-level rise that peaked during Mwp-1A being coeval with the collapse of the marine-based part of the Icelandic ice sheet suggests a causal relationship between the two events. Keywords Icelandic ice sheet Á Late Weichselian Á Ice sheet collapse Á Deglaciation Á Sea-level rise
Terrestrial and marine geological archives in the Arctic contain information on environmental cha... more Terrestrial and marine geological archives in the Arctic contain information on environmental change through Quaternary interglacialeglacial cycles. The Arctic Palaeoclimate and its Extremes (APEX) scientific network aims to better understand the magnitude and frequency of past Arctic climate variability , with focus on the " extreme " versus the " normal " conditions of the climate system. One important motivation for studying the amplitude of past natural environmental changes in the Arctic is to better understand the role of this region in a global perspective and provide base-line conditions against which to explore potential future changes in Arctic climate under scenarios of global warming. In this review we identify several areas that are distinct to the present programme and highlight some recent advances presented in this special issue concerning Arctic palaeo-records and natural variability, including spatial and temporal variability of the Greenland Ice Sheet, Arctic Ocean sediment stratigraphy, past ice shelves and marginal marine ice sheets, and the Cenozoic history of Arctic Ocean sea ice in general and Holocene oscillations in sea ice concentrations in particular. The combined sea ice data suggest that the seasonal Arctic sea ice cover was strongly reduced during most of the early Holocene and there appear to have been periods of ice free summers in the central Arctic Ocean. This has important consequences for our understanding of the recent trend of declining sea ice, and calls for further research on causal links between Arctic climate and sea ice.
Research campaigns over the last decade have yielded a growing stream of data that highlight the ... more Research campaigns over the last decade have yielded a growing stream of data that highlight the dynamic nature of Arctic cryosphere and climate change over a range of time scales. As a consequence, rather than seeing the Arctic as a near static environment in which large scale changes occur slowly, we now view the Arctic as a system that is typified by frequent, large and abrupt changes. The traditional focus on end members in the system e glacial versus interglacial periods e has been replaced by a new interest in understanding the patterns and causes of such dynamic change. Instead of interpreting changes almost exclusively as near linear responses to external forcing (e.g. orbitally-forced climate change), research is now concentrated on the importance of strong feedback mechanisms that in our palaeo-archives often border on chaotic behaviour. The last decade of research has revealed the importance of on-off switching of ice streams, strong feedbacks between sea level and ice sheets, spatial and temporal changes in ice shelves and perennial sea ice, as well as alterations in ice sheet dynamics caused by shifting centres of mass in multi-dome ice sheets. Recent advances in dating techniques and modelling have improved our understanding of leads and lags that exist in different Arctic systems, on their interactions and the driving mechanisms of change. Future Arctic research challenges include further emphases on rapid transitions and untangling the feedback mechanisms as well as the time scales they operate on.
While there are numerous hypotheses concerning glacialeinterglacial environmental and climatic re... more While there are numerous hypotheses concerning glacialeinterglacial environmental and climatic regime shifts in the Arctic Ocean, a holistic view on the Northern Hemisphere's late Quaternary ice-sheet extent and their impact on ocean and sea-ice dynamics remains to be established. Here we aim to provide a step in this direction by presenting an overview of Arctic Ocean glacial history, based on the present state-of-the-art knowledge gained from field work and chronological studies, and with a specific focus on ice-sheet extent and environmental conditions during the Last Glacial Maximum (LGM). The maximum Quaternary extension of ice sheets is discussed and compared to LGM. We bring together recent results from the circum-Arctic continental margins and the deep central basin; extent of ice sheets and ice streams bordering the Arctic Ocean as well as evidence for ice shelves extending into the central deep basin. Discrepancies between new results and published LGM ice-sheet reconstructions in the high Arctic are highlighted and outstanding questions are identified. Finally, we address the ability to simulate the Arctic Ocean ice sheet complexes and their dynamics, including ice streams and ice shelves, using presently available ice-sheet models. Our review shows that while we are able to firmly reject some of the earlier hypotheses formulated to describe Arctic Ocean glacial conditions, we still lack information from key areas to compile the holistic Arctic Ocean glacial history.
Reconstructions of the Late Quaternary glacial history of the Kara Sea area show repeated build-u... more Reconstructions of the Late Quaternary glacial history of the Kara Sea area show repeated build-up of ice-sheet domes over the shallow epicontinental Kara Sea. Inferred ice divides were situated over the central Kara Sea, and the ice sheet repeatedly inundated the surrounding coastal areas of western Siberia. Geological fingerprinting of the Kara Sea ice sheet include end moraine zones, raised beaches, tills, glaciotectonic deformations and coarsening-upward sediment sequences, reflecting isostatic rebound cycles. This paper reviews evidence from several areas along the perimeter of the Kara Sea, suggesting that peripheral sites were critical for the initiation of the large Kara Sea ice sheet. Ice-sheet inception progressed with the formation of local ice caps that later coalesced on the adjacent shelf with globally falling sea levels, eventually merging and growing into a large ice dome.
In times of chronic lack of resources for academic research and ever increasing competition for g... more In times of chronic lack of resources for academic research and ever increasing competition for grants it was every scientists dream coming true: a billionaire patron comes along and hands you unlimited resources to pursue the research that lies closest to your heart. In this case, the late Gary Comer (1927–2006), who had in 2001 taken his yacht through the notorious Northwest Passage then free of sea ice, engaged a team of outstanding climate scientists to lead a search for causal links controlling abrupt global climate change: Wallace (Wally) S. Broecker of Columbia University, George H. Denton of the University of Maine, and Richard B. Allen of Pennsylvania State University. Broecker is an oceanographer/geochemist, who is probably best known for developing the idea of a global " conveyor belt " linking the circulation of the global oceans and controlling large scale climate oscillations in the past; Denton is a geologist, long concerned with the (bi-polar) geological history of the large Quaternary ice sheets; and Alley is a glaciologist who is perhaps best known for his contributions on the relationships between Earth's cryosphere and global climate change. All three have developed ideas and concepts on rapid ocean-atmosphere-cryosphere reorganizations * Ólafur Ingólfsson earned his B.S.in geology from University of Iceland in 1979 and his Ph.D. in quaternary geology from Lund University, Sweden. He is on leave from the University of Iceland and is currently a professor of quaternary and glacial geology at UNIS. His research interests are the late-Quaternary glacial and climate histories of the polar regions, and has conducted field work in the Arctic (Iceland, Greenland, Svalbard, Western Siberia) every season over the past thirty years, as well as participated in five Antarctic expeditions. His Web of Science publication record contains sixty-nine publications in international peer-reviewed journals.
Uploads
Papers by Ólafur Ingólfsson