Periodic Reporting for period 1 - ESPSI (Eruption Source Parameters for Explosive Eruptions in Iceland Over the Last 3 ka)
Reporting period: 2017-06-15 to 2019-06-14
Iceland is one of the volcanically most active regions of the world, with over 20 eruptions per century and a majority (>70%) feature significant explosive activity. About 700 tephra-producing events are preserved in the post-glacial soil and lake sediments in Iceland and the portions representing the last 3 ka contain about 200–250 events. Eruption Source Parameters (ESPs)—namely eruption duration, erupted volume/mass, plume height, magma discharge rate, and total grain size distribution— have yet to be determined for most of these eruptions. ESPs are traditionally derived from field measurements via a range of empirical methods. A common feature of all these approaches is that they require detailed mapping of the tephra deposits, which is very time consuming and expensive in terms of man hours.
The principal objectives of this project were to (a) evaluate how effective inversion modelling of the tephra fallout from explosive eruptions is in estimating, to the first order, the range of ESPs for such event in Iceland and (b) correlate the tephra fall record, as preserved in lake sediment archives, in Western Iceland over the last 3 ka. The advantage of using inversion to reconstruct the ESPs for single eruptions is that it reduced the number of data points required for the reconstruction (i.e. saves time and cost) and thus bypasses the compilation of isopach maps, through optimization algorithms. Inversion is also very useful when there are not enough data points to construct precise isopach maps, but the dispersal axis is somewhat constrained.
This study focused on the tephra record over the last 3 ka from Hekla and Katla volcanoes. The chosen time period encompasses an estimated 160 eruptive events, thereof about 70 from Hekla and 90 from Katla. The targeted volcanoes are two of the three greatest tephra-producing volcanoes in Iceland. The post-3 ka activity at Hekla is a good representation of felsic (andesite to rhyolite) eruptions, while Katla activity is typified by explosive basaltic events. The aim of the project was to further the understanding of 1) the range of ESPs for the explosive eruptions produced by Hekla and Katla in the last 3 ka, 2) the ESPs for the most common explosive eruptions from Hekla and Katla, and 3) the ESPs of the most extreme events and how often they occur.
ESPs are key information for characterizing explosive eruptions, important input parameters in computer models used for real-time ash dispersal forecasting by the Volcanic Ash Advisory Centres and can also be used in hazard assessments. Hence, establishing a statistically robust data set of ESPs, for a small but representative subset of the explosive eruptions in Iceland is a critical task for both fundamental understanding of eruptive processes and for ash forecasting.
The principal objectives of this project were to (a) evaluate how effective inversion modelling of the tephra fallout from explosive eruptions is in estimating, to the first order, the range of ESPs for such event in Iceland and (b) correlate the tephra fall record, as preserved in lake sediment archives, in Western Iceland over the last 3 ka. The advantage of using inversion to reconstruct the ESPs for single eruptions is that it reduced the number of data points required for the reconstruction (i.e. saves time and cost) and thus bypasses the compilation of isopach maps, through optimization algorithms. Inversion is also very useful when there are not enough data points to construct precise isopach maps, but the dispersal axis is somewhat constrained.
This study focused on the tephra record over the last 3 ka from Hekla and Katla volcanoes. The chosen time period encompasses an estimated 160 eruptive events, thereof about 70 from Hekla and 90 from Katla. The targeted volcanoes are two of the three greatest tephra-producing volcanoes in Iceland. The post-3 ka activity at Hekla is a good representation of felsic (andesite to rhyolite) eruptions, while Katla activity is typified by explosive basaltic events. The aim of the project was to further the understanding of 1) the range of ESPs for the explosive eruptions produced by Hekla and Katla in the last 3 ka, 2) the ESPs for the most common explosive eruptions from Hekla and Katla, and 3) the ESPs of the most extreme events and how often they occur.
ESPs are key information for characterizing explosive eruptions, important input parameters in computer models used for real-time ash dispersal forecasting by the Volcanic Ash Advisory Centres and can also be used in hazard assessments. Hence, establishing a statistically robust data set of ESPs, for a small but representative subset of the explosive eruptions in Iceland is a critical task for both fundamental understanding of eruptive processes and for ash forecasting.
Two eruptions from Katla (in 1721 and 1918) and one eruption from Hekla (the so named Hekla Z eruption which took place somewhere between 550 and 950 BC) were chosen as the test case studies for this project. Thickness measurements and samples were collected across the footprint of each deposit and combined with previously published and unpublished datasets. The grain size distribution was analysed at 18 locations for Katla 1721, 37 locations for Katla 1918, and 7 locations for Hekla Z. The Katla 1721 and 1918 deposits are very fine grained and well sorted at most outcrops, consisting dominantly of ash (<2 mm in diameter). The Hekla Z deposit is considerable coarser and with <15% ash at the most proximal locations.
Deposit thickness was converted to tephra load (i.e. mass per area) using an average measured bulk deposit density of 920 kg m-3 for Katla 1721, 1010 kg m-3 for Katla 1918, and 520 kg m-3 for Hekla Z and used as the input for inversion modelling. Eruption Source Parameters for each eruption were estimated using the TEPHRA2 model in inversion mode, run on the High Performance Computing (HPC) computer cluster at University of Geneva. For Katla 1721, the erupted mass was estimated to be 2.1×1011–2.2×1011 kg, equivalent to a volume of 0.2–0.3 km3, with a plume height of about 19 km. For Katla 1918, the erupted mass was estimated to be 2.1×1011–3.9×1012 kg, equivalent to a volume of 0.2–3.9 km3, with a plume height in the range of 15–23 km. For Hekla Z, the erupted mass was estimated to be 6.8×1010–2.7×1011 kg, equivalent to a volume of 0.1–0.5 km3, with a plume height of about 16 km.
On the volcanic explosivity index, Katla 1721 was a VEI 4, Katla 1918 a 4–5, and Hekla Z a 4. On the magnitude scale, Katla 1721 was a M 4.3–4.5 Katla 1918 a 4.3–5.6 and Hekla Z a 3.8–4.4. On the intensity scale, Katla 1721 was a 10.6 Katla 1918 a 10.2–11.0 and Hekla Z a 10.3.
Deposit thickness was converted to tephra load (i.e. mass per area) using an average measured bulk deposit density of 920 kg m-3 for Katla 1721, 1010 kg m-3 for Katla 1918, and 520 kg m-3 for Hekla Z and used as the input for inversion modelling. Eruption Source Parameters for each eruption were estimated using the TEPHRA2 model in inversion mode, run on the High Performance Computing (HPC) computer cluster at University of Geneva. For Katla 1721, the erupted mass was estimated to be 2.1×1011–2.2×1011 kg, equivalent to a volume of 0.2–0.3 km3, with a plume height of about 19 km. For Katla 1918, the erupted mass was estimated to be 2.1×1011–3.9×1012 kg, equivalent to a volume of 0.2–3.9 km3, with a plume height in the range of 15–23 km. For Hekla Z, the erupted mass was estimated to be 6.8×1010–2.7×1011 kg, equivalent to a volume of 0.1–0.5 km3, with a plume height of about 16 km.
On the volcanic explosivity index, Katla 1721 was a VEI 4, Katla 1918 a 4–5, and Hekla Z a 4. On the magnitude scale, Katla 1721 was a M 4.3–4.5 Katla 1918 a 4.3–5.6 and Hekla Z a 3.8–4.4. On the intensity scale, Katla 1721 was a 10.6 Katla 1918 a 10.2–11.0 and Hekla Z a 10.3.
The work carried out during this project has increased our understanding of the size and frequency of eruptions from Hekla and Katla volcanoes in Iceland in the last 3000 years. It has also increased the available dataset of ESPs for eruptions in Iceland which will be useful in ash forecasting, hazard assessments and further studies dealing with the range of explosive activity from these volcanoes.