Esther Jensen - Academia.edu (original) (raw)
Papers by Esther Jensen
Jökull
An overview of registered avalanches in the vicinity of 13 towns known to be threatened by avalan... more An overview of registered avalanches in the vicinity of 13 towns known to be threatened by avalanches is presented graphically, together with information on weather prior to avalanches at specific locations. In N-Vestfirðir (NW-Iceland), Central N-Iceland and in Austfirðir (E-Iceland), large avalanches are generally preceded by heavy precipitation and strong sustained winds from northerly directions. In such cases, the snow accumulates at the top of the lee slopes. In some cases snow accumulates in gullies when the wind blows parallel to the mountain side and at some locations, snow accumulation is very sensitive to wind direction.
<p>Jökulhlaups from marginal and subglacial lakes are a considerable hazard in Ice... more <p>Jökulhlaups from marginal and subglacial lakes are a considerable hazard in Iceland and the rapid retreat of glaciers and ice caps is leading to hydrological changes in many locations at or near the glaciers. This calls for careful monitoring of glaciers and proglacial areas.</p><p>On August 17 2020, increased discharge was observed in Hvítá, a glacial river originating in the ice cap Langjökull. Sediment-laden jökulhlaup waters filled a narrow gorge of the river near the farm and tourist resort Húsafell and dead salmon were found strewn over fields 30–40 km downstream.</p><p>Reconnaissance trips, overflights and satellite image studies revealed the following course of events:</p><p>A marginal glacial lake (current size: 1.3 km<sup>2</sup>) started forming at 890 m elevation at the western margin of Langjökull after the turn of the century. Sentinel-2 satellite images indicate that subglacial outflow from the lake had started in the morning of August 17. The exact path of the 2 km long subglacial water course can be inferred from a Landsat-8 image taken on November 11 2020. The image shows a narrow surface depression resulting from lowering of the glacier surface when the subglacial tunnel carrying the water was formed. The ice thickness averages 70 m along the flowpath.</p><p>Emerging from beneath the ice cap, the water flowed 13 km through the Svartá river canyon, eroding sediment from the river bed and canyon walls. Fresh colouring and sediment deposition was observed on sandur plains where Svartá joins the Geitá and Hvítá rivers.</p><p>Observations of the jökulhlaup (water level and flow velocity) as it passed beneath a bridge near Húsafell help constrain discharge levels and flood volume at a location 18 km from the outlet at Langjökull. In addition, real-time data on Hvítá river water level are available from the Kljáfoss hydrometric station 35 km further downstream, discharge started rising from a background value of 90 m<sup>3</sup>/s on August 17 at 16:00. The flood peaked there at 260 m<sup>3</sup>/s at 01:45 in the early morning of August 18 and had subsided again at noon on that day.</p><p>Using imagery from the Sentinel-2 satellites the area of the marginal lake is estimated to have diminished from 1.29 km<sup>2</sup> to 0.46 km<sup>2</sup> during the jökulhlaup. A lowering of 4 m has been determined from aerial imagery and the total volume released was 3.4 million m<sup>3</sup> according to preliminary estimates. We estimate an average flow velocity of 3±1 m/s for the entire distance from the outlet at the glacier to Kljáfoss.</p><p>The glacier margin in the region has retreated by 500-1000 m and thinned by 3 m/a in the period 2004-2019 leading to the formation of the proglacial lake. Flooding events occurring in 2014 and 2017 have now been detected in hydrometric and remote sensing data. The lake is likely to become larger when retreat continues and further thinning of the ice may lead to more frequent jökulhlaups in coming years. Plans to monitor the lake level and install early warning systems will be outlined in the presentation.</p>
Hofundar: Esther Hliðar Jensen, Jorunn Harðardottir, Svava Bjork Þorlaksdottir, Snorri Zophoniass... more Hofundar: Esther Hliðar Jensen, Jorunn Harðardottir, Svava Bjork Þorlaksdottir, Snorri Zophoniasson, Sigriður Magnea Oskarsdottir
On the 20th of March 2007 a large rock avalanche fell on Morsárjökull, one of the outlet glaciers... more On the 20th of March 2007 a large rock avalanche fell on Morsárjökull, one of the outlet glaciers from the southern part of the Vatnajökull ice cap, in south Iceland. This is considered to be one of the largest rock avalanches which have occurred in Iceland during the last decades. It is believed that it fell in two separate stages, the main part fell on the 20th of March and the second and smaller one, on the 17th of April 2007. The Morsárjökull outlet glacier is about 4 km long and surrounded by up to 1000 m high valley slopes. The outlet glacier is fed by two ice falls which are partly disconnected to the main ice cap of Vatnajökull, which indicates that the glacier is mainly fed by ice avalanches. The rock avalanche fell on the eastern side of the uppermost part of the Morsárjökull outlet glacier and covered about 1/5 of the glacier surface, an area of about 720,000 m2. The scar of the rock avalanche is located on the north face of the headwall above the uppermost part of the gl...
&lt;p&gt;The Fagradalsfjall eruption started on the 19&lt;sup&gt;... more &lt;p&gt;The Fagradalsfjall eruption started on the 19&lt;sup&gt;th&lt;/sup&gt; of March 2021 on a ~180 m long eruptive fissure, following a dike intrusion which had been ongoing for approximately three weeks. The eruption focused shortly thereafter on two eruptive vents. In April, new fissure openings occurred northeast of the initial eruption on the 5&lt;sup&gt;th&lt;/sup&gt;, 6/7&lt;sup&gt;th&lt;/sup&gt;, 10&lt;sup&gt;th&lt;/sup&gt;, and 13&lt;sup&gt;th&lt;/sup&gt; of April. The northernmost eruption occurred on the 5&lt;sup&gt;th&lt;/sup&gt; of April, approximately 1 km northeast of the initial fissure, whereas the other fissure openings occurred between this and the initial eruptive vents. Stills from web cameras and time-lapse cameras are available for five of the fissure openings. These data show that the eruptions were preceded by steam emitted from cracks in the exact locations where the eruptions started. The time between the first steam observations and the visual appearance of glowing lava ranged between 15 seconds and 1.5 minutes during night observations and 9 to 23 minutes during daytime observations, the difference is likely explained by different lighting conditions. The eruptive vents are located where the north-easterly oriented dike intersected pre-existing north-south oriented strike-slip faults. These strike-slip faults could be identified on both pre-existing aerial photographs and digital elevation models. A high resolution ICEYE interferogram spanning the first day of the eruption in March reveals deformation where the later vent openings occurred in April. This indicates how Interferometric Synthetic Aperture Radar Analysis (InSAR) could be used to predict where subsequent vent openings are likely. This is of great importance for hazard assessment and defining exclusion zones during fissure eruptions.&lt;/p&gt;
&lt;p&gt;The Fagradalsfjall eruption started on the 19&lt;sup&gt;... more &lt;p&gt;The Fagradalsfjall eruption started on the 19&lt;sup&gt;th&lt;/sup&gt; of March 2021 on a ~180 m long eruptive fissure, following a dike intrusion which had been ongoing for approximately three weeks. The eruption focused shortly thereafter on two eruptive vents. In April, new fissure openings occurred northeast of the initial eruption on the 5&lt;sup&gt;th&lt;/sup&gt;, 6/7&lt;sup&gt;th&lt;/sup&gt;, 10&lt;sup&gt;th&lt;/sup&gt;, and 13&lt;sup&gt;th&lt;/sup&gt; of April. The northernmost eruption occurred on the 5&lt;sup&gt;th&lt;/sup&gt; of April, approximately 1 km northeast of the initial fissure, whereas the other fissure openings occurred between this and the initial eruptive vents. Stills from web cameras and time-lapse cameras are available for five of the fissure openings. These data show that the eruptions were preceded by steam emitted from cracks in the exact locations where the eruptions started. The time between the first steam observations and the visual appearance of glowing lava ranged between 15 seconds and 1.5 minutes during night observations and 9 to 23 minutes during daytime observations, the difference is likely explained by different lighting conditions. The eruptive vents are located where the north-easterly oriented dike intersected pre-existing north-south oriented strike-slip faults. These strike-slip faults could be identified on both pre-existing aerial photographs and digital elevation models. A high resolution ICEYE interferogram spanning the first day of the eruption in March reveals deformation where the later vent openings occurred in April. This indicates how Interferometric Synthetic Aperture Radar Analysis (InSAR) could be used to predict where subsequent vent openings are likely. This is of great importance for hazard assessment and defining exclusion zones during fissure eruptions.&lt;/p&gt;
Annals of Geophysics, 2018
During the emplacement of the 2014−2015 lava flow in Holuhraun (Iceland) a new code for the simul... more During the emplacement of the 2014−2015 lava flow in Holuhraun (Iceland) a new code for the simulation of lava flows (MrLavaLoba) was developed and tested. MrLavaLoba is a probabilistic code which derives the area likely to be inundated and the thickness of the final lava deposit. The flow field in Holuhraun progressed through a fairly flat floodplain, and the initial limited availability of topographic data was challenging, forcing us to develop specific modeling strategies. The development of the code, as well as simulation tests, continued after the end of the eruption, and latest results largely benefitted from the availability of improved topographic data. MrLavaLoba simu− lations of the Holuhraun scenario are compared with detailed observational analyses derived from the literature. The obtained results high− light that small−scale morphological features in the pre−emplacement topography can strongly influence the propagation of the flow. The distribution of the volume settling throughout the extension of the flow field turned out to be very important, and strongly affects the fit between the simulated and the real extent of the flow field. The performed analysis suggests that an improvement in the code should al− low adaptable calibration during the course of the eruption in order to mimic different emplacement styles in different phases.
In January 2020, inflation up to 5 cm was detected in the volcanic system of Svartsengi, Reykjane... more In January 2020, inflation up to 5 cm was detected in the volcanic system of Svartsengi, Reykjanes peninsula (Iceland). The inflation was probably linked to the movement of magma which was estimated to be at a depth of 3-5 km. Shortly after the detection of the inflation, the Scientific Advisory Board responsible for tackling the unrest deemed possible that the unrest could evolve into an effusive eruption. We used both the MrLavaLoba and the DOWNFLOW codes to simulate the area potentially inundated by lava flows in order to assess the hazard posed in case of an effusive eruption. The DOWNFLOW code was used to create a suite of 10,000 simulations which were used to derive maps of the lava flow hazards. These maps can be dynamically updated to account for ongoing modifications suggested by the geophysical signals of the monitoring system. The MrLavaLoba code, in turn, was tuned based on the historical lava flows in the area, so it would be ready to simulate potential lava flow fields...
On 14 April 2010 an eruption started in Eyjafjallajökull, in southern Iceland. This was an explos... more On 14 April 2010 an eruption started in Eyjafjallajökull, in southern Iceland. This was an explosive eruption in the caldera, beneath the glacier. During the first two weeks the eruption went through two phases, an explosive phase with much tephra and ash production and a calmer phase with less productivity and some lava production.
Jökull
An overview of registered avalanches in the vicinity of 13 towns known to be threatened by avalan... more An overview of registered avalanches in the vicinity of 13 towns known to be threatened by avalanches is presented graphically, together with information on weather prior to avalanches at specific locations. In N-Vestfirðir (NW-Iceland), Central N-Iceland and in Austfirðir (E-Iceland), large avalanches are generally preceded by heavy precipitation and strong sustained winds from northerly directions. In such cases, the snow accumulates at the top of the lee slopes. In some cases snow accumulates in gullies when the wind blows parallel to the mountain side and at some locations, snow accumulation is very sensitive to wind direction.
<p>Jökulhlaups from marginal and subglacial lakes are a considerable hazard in Ice... more <p>Jökulhlaups from marginal and subglacial lakes are a considerable hazard in Iceland and the rapid retreat of glaciers and ice caps is leading to hydrological changes in many locations at or near the glaciers. This calls for careful monitoring of glaciers and proglacial areas.</p><p>On August 17 2020, increased discharge was observed in Hvítá, a glacial river originating in the ice cap Langjökull. Sediment-laden jökulhlaup waters filled a narrow gorge of the river near the farm and tourist resort Húsafell and dead salmon were found strewn over fields 30–40 km downstream.</p><p>Reconnaissance trips, overflights and satellite image studies revealed the following course of events:</p><p>A marginal glacial lake (current size: 1.3 km<sup>2</sup>) started forming at 890 m elevation at the western margin of Langjökull after the turn of the century. Sentinel-2 satellite images indicate that subglacial outflow from the lake had started in the morning of August 17. The exact path of the 2 km long subglacial water course can be inferred from a Landsat-8 image taken on November 11 2020. The image shows a narrow surface depression resulting from lowering of the glacier surface when the subglacial tunnel carrying the water was formed. The ice thickness averages 70 m along the flowpath.</p><p>Emerging from beneath the ice cap, the water flowed 13 km through the Svartá river canyon, eroding sediment from the river bed and canyon walls. Fresh colouring and sediment deposition was observed on sandur plains where Svartá joins the Geitá and Hvítá rivers.</p><p>Observations of the jökulhlaup (water level and flow velocity) as it passed beneath a bridge near Húsafell help constrain discharge levels and flood volume at a location 18 km from the outlet at Langjökull. In addition, real-time data on Hvítá river water level are available from the Kljáfoss hydrometric station 35 km further downstream, discharge started rising from a background value of 90 m<sup>3</sup>/s on August 17 at 16:00. The flood peaked there at 260 m<sup>3</sup>/s at 01:45 in the early morning of August 18 and had subsided again at noon on that day.</p><p>Using imagery from the Sentinel-2 satellites the area of the marginal lake is estimated to have diminished from 1.29 km<sup>2</sup> to 0.46 km<sup>2</sup> during the jökulhlaup. A lowering of 4 m has been determined from aerial imagery and the total volume released was 3.4 million m<sup>3</sup> according to preliminary estimates. We estimate an average flow velocity of 3±1 m/s for the entire distance from the outlet at the glacier to Kljáfoss.</p><p>The glacier margin in the region has retreated by 500-1000 m and thinned by 3 m/a in the period 2004-2019 leading to the formation of the proglacial lake. Flooding events occurring in 2014 and 2017 have now been detected in hydrometric and remote sensing data. The lake is likely to become larger when retreat continues and further thinning of the ice may lead to more frequent jökulhlaups in coming years. Plans to monitor the lake level and install early warning systems will be outlined in the presentation.</p>
Hofundar: Esther Hliðar Jensen, Jorunn Harðardottir, Svava Bjork Þorlaksdottir, Snorri Zophoniass... more Hofundar: Esther Hliðar Jensen, Jorunn Harðardottir, Svava Bjork Þorlaksdottir, Snorri Zophoniasson, Sigriður Magnea Oskarsdottir
On the 20th of March 2007 a large rock avalanche fell on Morsárjökull, one of the outlet glaciers... more On the 20th of March 2007 a large rock avalanche fell on Morsárjökull, one of the outlet glaciers from the southern part of the Vatnajökull ice cap, in south Iceland. This is considered to be one of the largest rock avalanches which have occurred in Iceland during the last decades. It is believed that it fell in two separate stages, the main part fell on the 20th of March and the second and smaller one, on the 17th of April 2007. The Morsárjökull outlet glacier is about 4 km long and surrounded by up to 1000 m high valley slopes. The outlet glacier is fed by two ice falls which are partly disconnected to the main ice cap of Vatnajökull, which indicates that the glacier is mainly fed by ice avalanches. The rock avalanche fell on the eastern side of the uppermost part of the Morsárjökull outlet glacier and covered about 1/5 of the glacier surface, an area of about 720,000 m2. The scar of the rock avalanche is located on the north face of the headwall above the uppermost part of the gl...
&lt;p&gt;The Fagradalsfjall eruption started on the 19&lt;sup&gt;... more &lt;p&gt;The Fagradalsfjall eruption started on the 19&lt;sup&gt;th&lt;/sup&gt; of March 2021 on a ~180 m long eruptive fissure, following a dike intrusion which had been ongoing for approximately three weeks. The eruption focused shortly thereafter on two eruptive vents. In April, new fissure openings occurred northeast of the initial eruption on the 5&lt;sup&gt;th&lt;/sup&gt;, 6/7&lt;sup&gt;th&lt;/sup&gt;, 10&lt;sup&gt;th&lt;/sup&gt;, and 13&lt;sup&gt;th&lt;/sup&gt; of April. The northernmost eruption occurred on the 5&lt;sup&gt;th&lt;/sup&gt; of April, approximately 1 km northeast of the initial fissure, whereas the other fissure openings occurred between this and the initial eruptive vents. Stills from web cameras and time-lapse cameras are available for five of the fissure openings. These data show that the eruptions were preceded by steam emitted from cracks in the exact locations where the eruptions started. The time between the first steam observations and the visual appearance of glowing lava ranged between 15 seconds and 1.5 minutes during night observations and 9 to 23 minutes during daytime observations, the difference is likely explained by different lighting conditions. The eruptive vents are located where the north-easterly oriented dike intersected pre-existing north-south oriented strike-slip faults. These strike-slip faults could be identified on both pre-existing aerial photographs and digital elevation models. A high resolution ICEYE interferogram spanning the first day of the eruption in March reveals deformation where the later vent openings occurred in April. This indicates how Interferometric Synthetic Aperture Radar Analysis (InSAR) could be used to predict where subsequent vent openings are likely. This is of great importance for hazard assessment and defining exclusion zones during fissure eruptions.&lt;/p&gt;
&lt;p&gt;The Fagradalsfjall eruption started on the 19&lt;sup&gt;... more &lt;p&gt;The Fagradalsfjall eruption started on the 19&lt;sup&gt;th&lt;/sup&gt; of March 2021 on a ~180 m long eruptive fissure, following a dike intrusion which had been ongoing for approximately three weeks. The eruption focused shortly thereafter on two eruptive vents. In April, new fissure openings occurred northeast of the initial eruption on the 5&lt;sup&gt;th&lt;/sup&gt;, 6/7&lt;sup&gt;th&lt;/sup&gt;, 10&lt;sup&gt;th&lt;/sup&gt;, and 13&lt;sup&gt;th&lt;/sup&gt; of April. The northernmost eruption occurred on the 5&lt;sup&gt;th&lt;/sup&gt; of April, approximately 1 km northeast of the initial fissure, whereas the other fissure openings occurred between this and the initial eruptive vents. Stills from web cameras and time-lapse cameras are available for five of the fissure openings. These data show that the eruptions were preceded by steam emitted from cracks in the exact locations where the eruptions started. The time between the first steam observations and the visual appearance of glowing lava ranged between 15 seconds and 1.5 minutes during night observations and 9 to 23 minutes during daytime observations, the difference is likely explained by different lighting conditions. The eruptive vents are located where the north-easterly oriented dike intersected pre-existing north-south oriented strike-slip faults. These strike-slip faults could be identified on both pre-existing aerial photographs and digital elevation models. A high resolution ICEYE interferogram spanning the first day of the eruption in March reveals deformation where the later vent openings occurred in April. This indicates how Interferometric Synthetic Aperture Radar Analysis (InSAR) could be used to predict where subsequent vent openings are likely. This is of great importance for hazard assessment and defining exclusion zones during fissure eruptions.&lt;/p&gt;
Annals of Geophysics, 2018
During the emplacement of the 2014−2015 lava flow in Holuhraun (Iceland) a new code for the simul... more During the emplacement of the 2014−2015 lava flow in Holuhraun (Iceland) a new code for the simulation of lava flows (MrLavaLoba) was developed and tested. MrLavaLoba is a probabilistic code which derives the area likely to be inundated and the thickness of the final lava deposit. The flow field in Holuhraun progressed through a fairly flat floodplain, and the initial limited availability of topographic data was challenging, forcing us to develop specific modeling strategies. The development of the code, as well as simulation tests, continued after the end of the eruption, and latest results largely benefitted from the availability of improved topographic data. MrLavaLoba simu− lations of the Holuhraun scenario are compared with detailed observational analyses derived from the literature. The obtained results high− light that small−scale morphological features in the pre−emplacement topography can strongly influence the propagation of the flow. The distribution of the volume settling throughout the extension of the flow field turned out to be very important, and strongly affects the fit between the simulated and the real extent of the flow field. The performed analysis suggests that an improvement in the code should al− low adaptable calibration during the course of the eruption in order to mimic different emplacement styles in different phases.
In January 2020, inflation up to 5 cm was detected in the volcanic system of Svartsengi, Reykjane... more In January 2020, inflation up to 5 cm was detected in the volcanic system of Svartsengi, Reykjanes peninsula (Iceland). The inflation was probably linked to the movement of magma which was estimated to be at a depth of 3-5 km. Shortly after the detection of the inflation, the Scientific Advisory Board responsible for tackling the unrest deemed possible that the unrest could evolve into an effusive eruption. We used both the MrLavaLoba and the DOWNFLOW codes to simulate the area potentially inundated by lava flows in order to assess the hazard posed in case of an effusive eruption. The DOWNFLOW code was used to create a suite of 10,000 simulations which were used to derive maps of the lava flow hazards. These maps can be dynamically updated to account for ongoing modifications suggested by the geophysical signals of the monitoring system. The MrLavaLoba code, in turn, was tuned based on the historical lava flows in the area, so it would be ready to simulate potential lava flow fields...
On 14 April 2010 an eruption started in Eyjafjallajökull, in southern Iceland. This was an explos... more On 14 April 2010 an eruption started in Eyjafjallajökull, in southern Iceland. This was an explosive eruption in the caldera, beneath the glacier. During the first two weeks the eruption went through two phases, an explosive phase with much tephra and ash production and a calmer phase with less productivity and some lava production.