AURORA Intensive Field Course Iceland 2022

The study site of the AURORA Intensive Field Course was Seyðisfjörður, a village counting about 685 inhabitants that is located in the east of Iceland, at the head of the fjord Seyðisfjörður. The drive there from Reykjavík, where the participants met and the field course was officially started, took - with toilet and tourist breaks - a whole day.

Seyðisfjörður is prone to natural hazards, especially landslides and avalanches. In the last 140 years, 9 major landslide cycles occurred. The first extensive material damage dates back to around 1100 years ago [Illmer et al. 2016]. In general, most landslides are released in autumn, triggered by heavy precipitation.

An extreme precipitation event caused a massive landslide on 18 December 2020, being the largest landslide event ever recorded in Iceland. Many people lost their homes due to this event; thirteen houses and one museum were destroyed. Since then, lots of measurements have been taken to investigate the reasons of the landslide event and to enhance the housing security for people living in Seyðisfjörður.

Prior to the landslide event, heavy rainfall coincided with snow melt due to unusually warm temperatures for the given time of the year (December). The rainfall was unexpectedly persistent; it was forecasted to last for one day only, but the forecast remained the same for 5 consecutive days. The total precipitation for 7 days accumulated to 700 mm, which is extreme, especially when comparing it to the mean annual precipitation of Iceland’s capital Reykjavík (530 mm). Global climate change leads to more persistent low pressure systems in Iceland.

The heavy precipitation was also caused by northeasterly winds, which generally tend to bring more precipitation. People had been evacuated before the landslide event happened due to the likelihood caused by the longlasting rainfall.

For monitoring purposes, the site has been equipped with several devices. 45 reflectors are installed on the slope to trace their position from the opposite hill using a laser. This setting already revealed the annual slope movement to be around 1 cm. Temperature sensors were installed to stay informed about the soil condition. Several boreholes were drilled to place water level sensors inside. Inclinometers measure if the slope has already started to deform by measuring the angle. A ground-based InSAR takes one radar image per hour. These images can as well be used to monitor onsetting slope deformation. Four weather stations record the current weather parameters and store the measurements for later usage.

On the 16th of August we performed geophysical field measurements for possible Permafrost detection in areas prompt to future landslides up on Strandartindur mountain. Two ERT profiles were conducted and several GPR lines on, next to and perpendicular to the ERT lines. 

Ground penetrating radar allows us to use seismic refraction with radar pulses to image the subsurface. This involves sending a short electromagnetic pulse into the subsurface and then recording the energy reflected from the subsurface structure. The propagation and reflection of radar waves is controlled by three factors: permittivity, magnetic permeability and conductivity [Milsom, 2003].

Changes in these parameters affect the amount of energy reflected and therefore allow us to distinguish between different features, such as permafrost, ice, and faults (and many more) in the subsurface.

For conducting the measurements in the surroundings of Seyðisfjörður, the 100 MHz and 200 MHz antenna were used, to reach different depths and to get different resolutions close to the surface.

The GPR profile reveals some structures down to about 5 m. But more processing steps need to be done to hopefully achieve clearer images. There is a little deeper reaching, irregular, slightly bowl shaped reflection between 15 and 30 meters, that is approximately at the same location as the higher resistivity pocket in the ERT profile. This reflection could be from the interface between the permafrost pocket and the active layer above, but it is too early to draw too far-reaching conclusions from this. First we will need to find and assess the best processing steps and process all the rest of the GPR profiles.

ERT is sensitive to changes in physical properties of the subsurface and can produce 2- and even 3D tomographic models of the subsurface.

The measurement setup is a standard four-electrode setup where the currents flow perpendicular to the equipotential surfaces of the soil. The current is injected into two electrodes (B and A) and then the potential is measured between the other two electrodes (M and N).

This can be summarized in Ohm's law, where the apparent resistivity can be calculated as a function of the geometric factor K (depending on the measuring array) [Kearey et al., 2002, Krautblatter et al., 2010, Milsom, 2003].

The profile shows that we have a high resistivity pocket in the center of the profile (light blue pocket). Usually, the threshold for permafrost is 600 Ωm. Everything above that is considered permafrost (or frozen ground). Of course, this value can also be influenced by the rock properties itself. As this is just a first overview, we can consider this to possibly show permafrost. Nevertheless, the confidence of this interpretation is still quite low and further analysis needs to be done.

We are not sure yet how to interpret the high resistivity on the surface. We had quite a lot of gravel on the surface which could lead to higher resistivity. Note, that in our data set, the first measurement point was at around 0.8 m of depth so everything above this depth shown in the figure is interpolated in the inversion process.

Looking at publications for permafrost detection in the highlands in Iceland, pockets of permafrost (discontinuous) are quite frequent [Emmert & Kneisel, 2021].

The fourth day of the field trip was dedicated to Human Geography. We applied the method of participatory mapping in the context of natural hazards and residential area development by using the tool  Paper2GIS  [Huck et al. 201]. Participatory mapping or interaction with humans can be useful in the context of hazards research to gain insights into the attitude of the locals.

We split up into three groups of 3 to 4 people to interview local residents of Seyðisfjörður. The main goal was to ask them the following two questions and make them draw their answers on maps showing the town of Seyðisfjörður.

1. Where do you think residential areas should be developed in the future?
2. Where do you consider unsafe spaces with regards to natural hazards?

The interviewees were asked to draw their answers on the printed maps so that their drawn shapes could later be digitized automatically using Paper2GIS.

Paper2GIS extracts the shapefiles from the map drawings by remembering the initial state of the map without the drawings (Huck et al. 2017). The shapefiles can further be processed using any GIS software. In this case, QGIS was used to first create a hexagonal grid for the study area. In a next step, for each grid cell the number of polygons (extracted from the drawings) was counted. By adjusting the symbology a heatmap was created.

The AURORA Intensive Field Course ended the same way it had started: with a full day of driving. But again, the beautiful scenery compensated for the long journey back to Reykjavík.

 Huck, J.J., Dunning, I., Lee, P., Lowe T., Quek E., Weerasinghe, S. & Wintie, D. (2017). Paper2GIS: A self-digitising, paper-based PPGIS. In: Proceedings of the 14th International Conference on Geocomputation. 

 Illmer, D.; Helgasson, J.K.; Jóhannesson, T.; Gíslason; E.; Hauksson, S. (2016) Overview of landslide hazard and possible mitigation measures in the settlement southeast of Fjarðará River in Seyðisfjörður. Report VÍ 2016-006. 

Kearey, P., Brooks, M., and Hill, I. (2002). An Introduction to Geophysical Exploration. ISBN: 0-632-04929-4.

Kneisel, C., Hauck, C., Fortier, R. & Moorman, B. (2008), ‘Advances in geophysical methods for permafrost investigations’, Permafrost and Periglacial Processes 19(2), 157–178.

Krautblatter, M., Verleysdonk, S., Flores-Orozco, A., and Kemna, A. (2010). Temperature-calibrated imaging of seasonal changes in permafrost rock walls by quantitative electrical resistivity tomography (Zugspitze, German/Austrian Alps). J. Geophys. Res. Earth Surf., 115(2):1–15, ISSN: 21699011, DOI: 10.1029/2008JF001209.

Milsom, J. (2003). Field Geophysics. The Geological field guide series., volume 3. ISBN: 0470843470.

If not stated precisely by citing the relevant publications, all information in this report is based on what was explained during the field course by the assisting experts.