Field Workshop hosted by IAVCEI Working Group on Modeling Tephra-Fall Hazards and the Commission of Explosive Volcanism

Silicic explosive volcanism at divergent plate boundaries: Near-vent successions of a caldera forming eruption. The 1875 eruption of Askja volcano

12-17 August, 2008, Askja volcano, Iceland

by Alain Volentik, University of South Florida

Rebecca Carey and Bruce Houghton from the University of Hawai'i organized this intensive, six-day short course on Askja volcano within the North Volcanic Zone in Iceland. The workshop was hosted by the IAVCEI Working Group on Modeling Tephra-Fall Hazards and the Commission of Explosive Volcanism and took place during 12-17 August 2008, immediately before the 2008 IAVCEI General Assembly held in Reykjavik, Iceland. A total of thirty-one participants (including the two organizers) form various nationalities and disciplines converged in Northern Iceland to attend this workshop. The goal of the workshop was to use the 1875 eruption of Askja to address and discuss the following themes: (i) Enhanced proximal deposition and of its implications for eruption plume dynamics, (ii) Process of welding during powerful eruptions and the implications for eruption dynamics, (iii) Modeling wet plumes and (iv) Investigation of quantitative field methods, such as grain size and MP-ML measurements.

Day 1 (August 12): Workshop participants gathered in Akureyri in Northern Iceland and once the group was ready, and after stopping at the supermarket to buy provisions for the next 5 days, we headed off for the long drive (about 6 hours by bus) toward Askja volcano. We arrived quite late at our base camp, got settled in the nice cabin and had a late group dinner before catching up some rest.

Day 2 (August 13): After an early breakfast, the workshop leaders gave us an introduction on the 1875 eruption of Askja volcano and the resulting pyroclastic deposits. Participants were introduced to the general pyroclastic stratigraphy of the eruption, including the dry subplinian B tephra unit, the wet phreatoplinian C1 unit, the pyroclastic density current C2 unit and the dry Plinian D unit. We then headed toward the Askja car park and hiked toward the Oskjuvatn caldera which was produced during- and for approximately 40 years after the eruption and now hosts Iceland’s deepest lake. Our first stop was at the Viti crater (Figure 1), a post-main phase phreatic cone to the north of Oskjuvatn, where the leaders gave a brief overview of the geography of Askja and more details of the upper tephra sequence (Plinian unit D) of the 1875 eruption of Askja.


The Viti

Figure 1: The Viti crater, with lake Oskjuvatn (3 km x 4 km) in the top left of the picture.

Field activities that day focused mainly on the subdivision within the D unit: D1, D2, D3, D4 and D5. The second stop was located just east of the Viti crater on a peninsula where the internal stratigraphy of the D unit was particularly clear and accessible (Figure 2). Workshop participants were able to make critical observations of the internal components of each D sub-unit and the nature of the transitions between sub-units. Sub-units D1, D3 and D5 are white in color and coarse-grained, while sub-unit D2 and D4 are dark brown in color, poorly sorted with a substantial increase in matrix abundance, comprising predominantly fluidal juvenile ash. Some occasional spatter bombs were observed within the D4 sub-unit. D1, D3 and D5 cannot be distinguished from one another when the dark brown D2 and D4 sub-units are absent. These characteristics triggered interesting discussions on the plume and vents dynamics at the time of emplacement of the D tephra unit. Basically, the D2 and D4 sub-units thin very quickly away from the vent and have therefore a limited local dispersal. D1, D3 and D5 are regionally dispersed with Plinian-type thinning-half distances. From the dispersal data shown, sub-units D2 and D4 had different vents than that of D1, D3 and D5. Furthermore the nature of the D2 and D4 versus the D1, D3, D5 sub-units made it clear that the eruption dynamics were different. The D2 and D4 layers were produced by fountaining vent(s) that were erupting synchronously with the main Plinian plume responsible for the sedimentation of the D1, D3 and D5 layers.

 Fig2The D tephra unit

Figure 2: The D tephra unit. Note the D1, D3 and D5 light-colored sub-units and the D2 and D4 dark brown sub-units (Picture by Rebecca Carey).

The next stop featured a rhyolitic fountain-fed lava and dyke, which are precursory events to the main 1875 eruption (B-C-D). The rhyolitic lava flow displayed a basal obsidian-like layer developed upon flow emplacement. These products overlain, and cut through respectively, an older phreatomagmatic tuff cone. On our way back toward the Viti crater along the lake shore further precursory deposits were observed: a basaltic tuff ring cut by a silicic dyke (Figure 3).

Fig3Precursory activity to the main 1875
        explosive eruption of Askja volcano

Figure 3: Precursory activity to the main 1875 explosive eruption of Askja volcano. The silicic dyke cuts through the basaltic tuff cone tephra (Picture by Martin Jutzeler).

Day 3 (August 14): Field activities on the third day of the workshop focused mainly on welding deposits and processes associated with phase D. Participants also had the opportunity to make observations and discuss the Phreatoplinian C1 tephra and C2 pdc's. Intense and interesting discussions on the welding processes occurred due to first-time observations for most of the group of welding calc-alkaline silicic fall, in addition to the outstanding nature of the outcrops and also the complexity of two separate welding processes – regional welding: due to the intrinsic heat of the deposit and local welding: due to both load and heat dissipation from spatter bombs into adjacent deposits (Figure 4b). The welding is related to the above mentioned D2 and D4 fountaining deposits, which weld the individual sub-units in addition to encroaching upon the Plinian D1, D3 and D5 sub-units. At many stages along the northern rim, the non-welded thickness of the D deposits can be severely reduced when there is a very high degree of welding, often producing a completely welded unit which is densely welded. The workshop leaders also introduced their work on the spatter bomb size and geographical distribution toward the north away from the caldera. From their distribution, a tri-lobate-shaped pattern was defined that suggests that the fountains were directional at times during fountaining activity, and this distribution also supported their interpretation of multiple fountaining vents during D2 and D4 times based on regional welding patterns explained earlier during the day.

Welding horizon
        within the Askja D unitFig4a

Figure 4a: Welding horizon within the Askja D unit, which is encroaching upon sub-unit D3, which appears black in this photo and welded to a lesser degree.

Fig4bPart of the group discussing field

Figure 4b: Part of the group discussing field observations on welding processes in the Askja D unit.

On the final stop of the day, participants reached a beautiful outcrop displaying pyroclastic density current deposits (C2). Interesting discussions on bedforms, relative energy and water involvement took place at that location, together with the determination of their likely source region based on flow directions. Beneath the surge sequences lies the Phreatoplinian deposit, and participants dug a huge hole to reach the bottom of the Askja C unit in order to see the thickness (Figure 5). Discussions on the Phreatoplinian C1deposit (Figure 5) focused mainly on how a wet phreatoplinian unit can be separated from highly and efficiently fragmented dry deposit and how these deposits can be modeled (or not) with current tephra fallout models. Eruption column heights for dry plumes are calculated using MP and especially ML data, but in the case of highly fragmented wet eruptions, such as the Askja C unit, these models cannot be applied and therefore other approaches have to be developed. Furthermore, participants discussed potential processes that control the fall of such fine ash so close from the vent, while current tephra models based on advection-diffusion approaches, would predict the sedimentation of such fine grained material far from the vent. Is the high density of the wet eruption column promoting premature fallout of fine ash? Are there dilute density currents depositing material close to vent? Are the particles falling out as wet aggregates? These processes will have to be fully understood in order to build tephra dispersal model that account for wet eruptions such as the Askja C tephra unit.

Fig5Askja C unit

Figure 5: Askja C unit, with the Phreatoplinian tephra fall at the base (overlaying the pre-1875 ice sheet) overlaid by the surge sequences. Dr. Rebecca Carey sampling the pre-1875 ice for the end-of-the-day whiskey on the rocks.

Day 4 (August 15): The day started with the long hike from the Askja car parking lot toward the western side of the caldera rim. Our first stop focused on a black, very poorly sorted deposit lying above the C2 pdc's. The nature of this pyroclastic deposit was foreign to what the group had seen in the last few days. We then crossed an a'a basaltic lava and reached the southwestern side of the caldera, where we once more observed the poorly sorted unit, but it transitioned to an upper red color and also had distinct bedforms. Therefore, the deposit was interpreted as a pdc deposit. The pdc's were also definitely sitting above the Askja unit C2 pdc's and toward the south, dense spatter bombs were present in discrete layers within these deposits similar to those observed on the northern rim, in the Askja D unit. While walking toward the south, the abundance and size of the spatter bombs increased, the red pdc's thickened and welding started to appear and intensify toward the south. Further to the south, the welding became more intense until we ultimately observed rheomorphic welded deposits. The rheomorphic deposits have been folded and formed roll structures along the southwestern shore of the lake (Figure 6). These rolls represented a rare opportunity to observe 3D characteristics of rheomorphic welded fall deposits, as they showed both cross-sectional and along-axis outcrops, together with internal features. The axis of these rolls is presently found parallel to the slope, which gave an indication on the rolls formation and emplacement. First, the rheomorphic deposits started to form folds perpendicular to slope, once formed, these folds started to be deformed in a V-shaped pattern downslope. This interpretation was confirmed by the finding of a preserved hinge in the rheomorphic rolls. A very interesting day!

Fig6Rheomorphic rolls on the southwestern
        side of the Askja caldera

Figure 6: Rheomorphic rolls on the southwestern side of the Askja caldera. Note their orientation parallel to slope, and their brittle fracture perpendicular to slope.

Day 5 (August 16): Today's activities took place at only one location for an exercise on grain size distribution and measurements of the Maximum Pumice (MP) and Lithic (ML) sizes. After a short briefing at our base camp, workshop participants were subdivided into 6 different groups, assigned the exercise materials (Figure 7) and then proceeded to the eastern shore of Oskjuvatn lake, location of the exercise. The exercise consisted for each group in (1) getting a grain size distribution (Figure 8) and MP and ML measurements by channel sampling through the whole D tephra unit (3.3 meters vertical stratigraphy) using a mass of total sample that the group collectively thought was appropriate, and (2) getting a grain size distribution and MP and ML measurements of a distinct sub-layer of the D tephra unit, which was defined by the workshop leaders and with their strategy for sample mass collection (largest clast must not weigh more than 5% of the total sample weight). The goal of such an exercise was (1) to compare the results in channel sample grain size distribution and maximum clast size measurements between the different groups, test what sampling methods trained volcanologists typically use for grain size distribution in the field, and (2) to see the variation in grain size distributions and maximum clast sizes within the D unit by analyzing the different sub-layers. The results of our exercise have been presented at the Tephra Group Meeting during the 2008 IAVCEI General Assembly in Reykjavik and the main result can be summarized as follows: First, the results of the grain size distribution analysis on the whole D tephra unit is not consistent between the different groups, as each group applied its own approach to the exercise. There will be therefore a need in the near future to standardize the approach to grain size distribution measurements in the field among volcanologists. This is extremely necessary as comparisons need to be made between research groups and these data need to be used in order for the eruption dynamics/plume modeling community to generate and provide as accurate tephra dispersal and sedimentation models as possible. Second, channel samples are not acceptable to characterize the grain size of a fallout unit, when obvious stratigraphy is present.

Fig7Material used for the exercise on
        grain size distribution and maximum clast size measurements

Figure 7: Material used for the exercise on grain size distribution and maximum clast size measurements (Picture by Stefen Kutterolf).

performing the
        grain size distribution exerciseFig8

Figure 8: Yasuo Miyabuchi, Corrado Cimarelli and Licia Costantini performing the grain size distribution exercise (Picture by Mannen Kazutaka).

Day 6 (August 17): We left our base camp early in the morning. On our way back to Akureyri, we stopped at Myvatn for welcomed 1-hour long outdoor hot spring session (Figure 9). All the participants were then dropped off in Akureyri in early afternoon for their trip back to Reykjavik to attend the 2008 IAVCEI General Assembly.

enjoying the hot
        spring in MyvatnFig9

Figure 9: Part of the group enjoying the hot spring in Myvatn (Picture by Mannen Kazutaka).

In conclusion, this workshop provided a unique opportunity to study and discuss proximal tephra sedimentation, plume dynamics, vent migration, transitions between wet and dry eruptive phases in addition to welding and rheomorphic processes. The workshop was a total success thanks to the exceptional organization and great enthusiasm of the leaders and the “sun-dances” that they performed in order to get us some good weather. Participants had the opportunity to study in detail beautifully exposed sections of rare proximal tephra deposits, share ideas and opinion on conduit, eruption and plume dynamics, and hopefully future scientific collaborations will emerge from this successful workshop on Askja volcano.

GroupGroup picture of the 2008
        workshop on Askja volcano

Group picture of the 2008 workshop on Askja volcano. Participants are standing on a huge spatter bomb (9 x 7 x 1.8 meters) associated with sub-unit D4. List of participants (in alphabetical order): Fabrizio Alfano, Sébastien Biass, Costanza Bonadonna, Michael Branney, Kazutaka Mannen, Shinji Takarada, Rebecca Carey, Corrado Cimarelli, Licia Costantini, Wim Degruyter, Judy Fierstein, Armin Freudt, Chris Gregg, Wes Hildreth, Bruce Houghton, Martin Jutzeler, Mannen Kazutaka, Laura Kerber, Malin Klawonn, Takehiro Koyaguchi, Stefen Kutterolf, Jessica Larsen, Jocelyn McPhie, Yasuo Miyabuchi, Danilo Palladino, Wendy Perez Fernandez, William Rose, Tanya Shavalia, Sigurjon Thorarinsson, Kae Tsunematsu and Alain Volentik.

(Last update Jan. 12, 2009)