After DISC lab in Bonn, Andreas Kemna drove us to the Cologne/Bonn airport where we caught an evening flight to Vienna. This was our most intensely-scheduled week for the European leg of the DISC (Bonn on Monday and Tuesday, Veinna on Thursday and Friday).
We landed in Vienna around 9:30 in the evening and headed straight to our airbnb accommodation. In typical Viennese style, the building had 4 floors (note: they use 0-based indexing to count floors in many parts of Europe, including Vienna!). We walked through the large wooden doors at the entrance, through the foyer, to the elevator. It was an old-style elevator, with a steel mesh door you open and a wooden door you slide back to enter; it was surrounded by a spiral staircase.We took the elevator up to the top floor and got settled; we were grateful to have a day to prep / rest ahead of us.
On Wednesday, we went to the Geologic Survey of Austria to see the room and test the equipment. We were very impressed by the level of professionalism; there were 3 people waiting to assist us and help set up the room for course. The Geologic Survey of Austria was by-far the best venue we have presented at. The projector was very high-quality, each participant had access to power for their laptops during the day, and name-cards for each participant were provided. Importantly, the coffee served during the breaks was fantastic. As a group at UBC, we have invested a lot of effort in putting together very visual slides — it is exciting to give a presentation in a setting that helps that work shine.
Day 1: DISC Presentation
The vast majority of the participants in Vienna were with the Geologic Survey; participants spanned a number of groups within the geologic survey, including: mineral exploration, oil and gas, engineering geology and natural hazards, hydrogeology, geomagnetic hazards. There were also a few attendees from industry, including in geoscientific software and geotechnical engineering.
At each location, we aim to present material and case histories that are relevant to local geoscientists — input from the local hosts is essential for making this happen. Robert Supper, the director of the Applied Geoscience Department (and our local host!), provided us with a case history on landslides in of Austria, a problem of relevance throughout the country. The case history is based on the 2013 paper Airborne geophysical mapping as an innovative methodology for landslide investigation: evaluation of results from the Gschliefgraben landslide, Austria by Supper et al.
The problematic unit in the Gschliefgraben area is a clay layer: the Buntmergelserie unit. When it rains, this layer absorbs water and becomes a plane of weakness, which can result in a landslide. This unit is more electrically conductive than its neighbouring geologic units, so an airborne electromagnetic survey was included in the surveys selected to map the clay layer. The airborne frequency domain system was developed by the Geologic Survey of Austria. One interesting note on the survey — due to the significant topography of the region, data were only collected while the helicopter was ascending, so the survey required twice as much flight-time as would have been required for flat topography. The data were inverted in 1D and the problematic clay unit delineated in the model. This interpretation was then incorporated into the landslide model for the area and used to inform the design of a landslide early warning network in the region. This case history is discussed in the slides on Inductive Sources.
From the participants
“Always inspiring to listen to experts, learning what is possible with EM. Long day, many slides, hard to remember [it] all.”
“It was a very informative day! We covered tonnes of material, but my favourite [aspect] was the way Doug related each method to the others. It could easily have been a 2 day course + DISC lab (3 days)”
“Very informative. It gave a good overview of the methods, both theory and practical applications. I really appreciated the use of applications to understand the physics of the method”
“It was intense but very informative. It provided one with a new insight into geophysics and my understanding of geophysics (even though limited) has improved.” — Sedimentologist with the Geological Survey of Austria
Day 2: DISC Lab
DISC Lab day started at 9am and we jumped right into discussion and presentation from participants. About half of the participants from day 1 returned for day 2, and nearly everyone who attended contributed a presentation. We were impressed by the quality of the presentations and conversations.
Robert Supper started off the DISC Lab conversations on the whiteboard. For a number of years, the Geologic Survey of Austria has been involved in a groundwater project in Mexico. In the Yucatan, much of the groundwater flow if controlled by Karst structures. Since water tends to be more electrically conductive than the rock matrix surrounding it, electrical and EM methods are of interest for mapping the Karst channels. Over one test-site, 3 boreholes have been drilled and several geophysical experiments have been conducted. A DC resistivity experiment has been conducted on the surface (often called geoelectrics in Europe), a surface-to-borehole electromagnetic experiment has been run, and borehole induction measurements have been made. The puzzle that they have encountered is that the borehole logging measurements are completely different from the geoelectrics… Why? Could it be because each is sensitive to different scales: is the borehole logging primarily sensitive to the pore conductivity while the DC and borehole to surface EM is sensitive to the bulk conductivity (rock matrix + pore fluid)? If so, could we combine surface measurements with borehole measurements to distinguish between porosity and connected flow paths?
Arnulf Schiller (slides) followed Robert and continued the discussion on Karst characterization in Mexico focussing on airborne EM data that have been collected over the Tulum Karst plain. The aim of this work is to develop a hydrologic model and understand the freshwater resource, both in terms of capacity and dynamics, in the area. Karst tunnels are high-permeability pathways for water to flow, so their location is an essential piece of information for building up a hydrological model.
Tulum is approximately 250km south of Cancun. The limestone unit in which the karsting occurs is several hundred metres thick, extending to a kilometre in some regions. There is minimal topography; it is a very flat region about 50m above sea level, but it is difficult to access by land, making it an ideal candidate for airborne surveys. It is expected that the Karst tunnels will be filled with water, saline seawater below the halocline and freshwater, which is still relatively electrically conductive (~4 Ωm), above it, making electromagnetics a potentially diagnostic method. Several hundred kilometres of the Karst tunnels have been explored by divers, providing some ground-truth. The questions to be addressed are:
- Can airborne EM be use to map the Karst conduit system?
- Can a Karst water regime be modelled by combining airborne geophysical data and hydrological input data?
Three airborne EM surveys have been flown with the Austrian system (a 4 frequency FDEM system), one in each of 2007, 2008 and 2015. The surveys cover the inland region and have some data just off-shore over the reefs. Arnulf showed data from the 2007 and 2008 surveys. Apparent resistivity maps were created from the raw data and several known caves were clearly outlined providing confidence that karsting can be imaged with the airborne EM data. The next step was to perform 1D inversions of the airborne EM data to estimate an electrical conductivity model of the region. With the first-pass inversion results, a significant conductivity contrast, which was interpreted to be the halocline was imaged. From this, a map of the depth to the interface between saline and fresh groundwater was created.
In order to try and extract more from the data, they applied further signal processing and inverted the processed data. From these inversion results, they were able to resolve the Karst conduit network. The groundwater level remains a challenge and is not well-resolved in the current inversion results. As a potential next step, we discussed mechanisms by which a-priori information can be included in an inversion scheme to improve the results (possibly using SimPEG!). For example, if known, bounds can be put on the resistivity of various units or abrupt changes in the conductivity model can be accommodated (rather than smoothed-over) if the location of geologic boundaries is known.
Arnulf also mentioned that there is interest in imaging the reef and lagoon area. Several lines were flown over the reef and a resistive structure was detected beneath seawater. Could airborne EM be used to monitor the coral reefs?
Ingrid Schattauer presented work she has done in an modelling and inversion study. There are many choices to be made when setting up an inversion:
- selecting an appropriate discretization,
- choosing a starting model and a reference model,
- defining a stopping criteria,
- assigning uncertainties on the data,
- choosing a trade-off parameter to weight the relative influence of the data misfit and regularizations and selecting if that is a fixed parameter in the inversion or if it is “cooled” during the inversion, progressively reducing the influence of the regularization as the inversion progresses (called a beta-cooling scheme), …
She compared the impact of a number of these parameter choices on the stability of the inversion and the character of the recovered model for the 1D inversion of airborne EM data. In the choice of trade-off parameter, she found that selecting a fixed trade-off parameter tended to be more stable than a beta-cooling approach. For the discretization, choosing a large number of layers and relying on the regularization rather than selecting a small number of layers (and imposing assumptions in the inversion through the discretization) also resulted in a more stable inversion. Doug re-iterated these conclusions stating that they are in-line with much of the experience at UBC. Ingrid’s work was conducted for 1D inversions, treating each sounding independently, a potential next-step would be to employ regularization between adjacent soundings so that smoothing is not only applied vertically, but laterally as well.
Ingrid then showed a field example from Tyrol, Austria. Gypsum is present throughout the region and makes the area susceptible to sink holes as gypsum is easily dissolved. In order to understand the hazard sinkholes pose in an area, it is important to have a good groundwater model. Airborne EM was flown over the region and the aim is to invert those data to identify karst regions and build up a groundwater model of the area.
Birgit Jochum (slides) started off her presentation discussing the DC / IP system that has been developed by the Geologic Survey of Austria. Several practical questions arise with respect to the cables which connect the electrodes. If the cable between each electrode is 3m long but a 1m electrode spacing is used for the survey, does the cable layout matter? is there a difference in the signal if it is looped vs not? maybe it should be looped in a figure-8 to cancel the magnetic flux from each loop? There is also the suggestion that shielded cables could be used: what are the advantages? disadvantages? Shielded cables are much heavier… The topic of shielded cables has come up in a number of conversations we have had recently; it is not a topic we have much background in, but something it seems we should be looking in to!
One application of DC that Birgit is working on is landslides. Understanding the how water infiltrates into the earth and where preferential flow paths are are important pieces of information for characterizing a landslide hazard. In a soil-substrate, when water infiltrates into the earth, the electrical resistivity is reduced, so DC resistivity can be an effective monitoring method. Birgit demonstrated time-lapse inversions where infiltration after a rainfall was imaged. The challenge is, most landslide hazards in Austria are clay landslides. Clay, even when dry, is quite conductive, so it is very challenging to observe changes in electrical conductivity due to water infiltration. Perhaps IP has some potential? It could still be quite a challenge as clay is also chargeable; changes due to water infiltration might be subtle… something to look in to.
Rachel Bailey gave her presentation on geomagnetically induced currents and the hazard they pose to the power grid (which we also encountered as an application of interest at the DISC Lab in Denver). During a geomagnetic storm, charged particles are ejected from the sun and when they hit the earth’s magnetic field, they cause very rapid variations in the geomagnetic field. This in-turn induces geoelectrical currents into the conductive subsurface and these can the be transferred to the power-grid. If too much current is loaded onto a power grid unexpectedly, it can cause transformers to fail, like during the 1989 blackout in Quebec.
To assess the hazard that geomagnetic storms poses, Rachel and her colleagues are interested in being able to model the currents induced into the earth and subsequently transferred to the power grid during a geomagnetic storm. In order to model the currents that are induced into the ground, a conductivity model of the subsurface of Austria is needed. For this, she is using the European Rho Model, an electrical conductivity model of continental Europe developed by the Research Centre for Astronomy and Earth Sciences in Hungary. By running a numerical simulation, the electric fields that would be observed at the surface of the earth during a storm can be estimated. These values are then used as input into a circuit model of Austria’s power grid which is subsequently solved, providing an estimate of the currents loaded into the power grid. Rachel demonstrated the importance of using a good conductivity model of the subsurface by comparing observed currents in the power grid with those predicted when using 3 different conductivity models of Austria, each with increasing spatial resolution. In addition to comparing the currents observed in the power grid with simulated results, they would also like to compare values of the electric field at the surface of the earth near the power stations. However, it is difficult to make such measurements as power stations produce a significant amount of electromagnetic noise…
Florian Bleibinhaus (slides), who is the head of the applied geophysics group at Montanuniversitaet Leoben, provided an overview of a number of problems that the geophysics group has been involved in. Florian’s background is primarily in controlled-source seismics, but has been involved in projects using DC resistivity as well as GPR. There are many similarities between the processing and interpretation of both GPR and seismic as both are governed by the physics of wave propagation.
The first example Florian showed was the application of DC and GPR for a hydrogeologic characterization project. There was interest in constructing a hydroelectric power plant along a river and in order to do so, the structure of the aquifer needed to be understood. The conductivity model from DC resistivity was combined with the GPR data to interpret the structure of the sand aquifer, and overlying silt and sand layers. Next, he showed several examples of the use of DC and GPR for archeological prospection. The final example Florian showed was using DC resistivity to monitor the stability of the slope adjacent to a gravel pit. Drainages were built as an effort to mitigate mass movement and they wanted to see how efficient those were for draining water. A time-lapse DC experiment was conducted with the intent of monitoring changes in resistivity after a rainfall event. Unfortunately, while the DC experiment was deployed, there was very little rainfall, so there was not much signal to work with.
Heinz Reitner (slides) gave a presentation on the application of DC resistivity for wine-growing. The Carnuntum region of Austria is a a wine-growing area, and the wine growers in this area are interested in understanding how the properties of the soil affect the grapes. A multi-disciplinary team, including soil scientists, geochemists and geophysicists were involved in the 3 year study. A variety of data sets have been collected and made available through a web application, including:
- microclimate observations: eg. precipitation, air & soil temperature, soil moisture among others
- soil: type, parent rock, texture, hydraulic properties
- geology: lithologic mapping, rock sample analysis
- geochem and soil chem: analysis of main and trace elements
- geophysics: apparent resistivity from geoelectrics.
The landscape is covered in growth, meaning no outcrops and limited soil-exposure, thus geophysics is important for providing distributed information about the soil throughout the vineyard. Electrical conductivity is an important indicator for moisture and salinity of soils, thus DC resistivity (also called geoelectrics) can provide valuable information for the wine-growers. To conduct the survey, the geocarta system was used. The geocarta system is a cart with 4 pairs of spikey metallic wheels that can be towed behind an ATV (going as fast as 20km/h!). Each spike acts as a pole in the DC resistivity survey giving 10cm along-line resolution. One wheel pair is used to inject current and the 3 other wheel pairs, each having a different separation, measure potential differences. An apparent resistivity map is made for each of the 3 receiver wheel pairs, and as they have different separations, each is sensitive to different depths (~50cm, ~100cm, ~200cm). For an image of the system, see slide 7 of Heinz’ slides.
Heinz showed examples of data that have been collected. Wine growers are using these data for “precision viticulture.” By knowing which regions of their vineyard have more gravelly vs more clayey soils, they can make more informed decisions about irrigation, the application of fertilizers, and harvesting. Since all of the data have been made available online, wine growers can access data and maps on their phone while they are out in the field.
The moral of the story: People want to drink good wine. Geophysics can help.
Sara Lise Underhay (slides), who works with Geosoft, provided an overview of where EM is used within Geosoft and focussed in on the particular application of Time Domain Electromangetics (TDEM) for locating and classifying Unexploded Ordnance (UXO).
We had thought that in many of the locations we have visited (throughout Europe, as well as in Korea, and Japan) that locating and classifying UXOs would be a problem of interest to many, but we have been surprised to see that it is not an application that particularly connects with many of participants. Perhaps is because it is more closely connected with the military rather than geoscientists in many countries? or liability and insurance requirements restrict who can be involved?
The majority of UXO surveys conducted are land-based. In the past, many were conducted with an EM-61 system, which is a dual loop system having one transmitter loop and one receiver loop; measurements are taken at a limited number of time-gates. It is essentially a large metal-detector; it is suitable for detecting UXO, but also detects other metallic scraps, and without more information, it is very challenging to tell the difference between scrap metal and an ordnance object. Even the same object, but at different orientations will give a different response.
One of the biggest costs in cleaning up an ordnance site is due to digging up scrap metal. In order to classifying an item as UXO or not prior to digging it up requires more data be collected. One system that has been developed for this purpose is the Geometrics Metal Mapper. Geosoft has been working with Geometrics on a US Department of Energy project to detect and classify UXO using the Metal Mapper. The survey is conducted in two steps. In the first step, a reconnaissance survey is used to identify potential targets and in the second step, the Metal Mapper is parked over the target item and multiple transmitters, with different orientations, excite the target from multiple directions, giving data that are sensitive to the orientation and depth of the ordnance object. To classify objects, a library of objects and their associated polarizibility curves is built. Measured data are then compared to this library (allowing for some variability, if for example, an item is damaged), and a dig / no dig list is created. Not all items can be cleanly classified, for those, they are dug up and added to the catalog of items, so that if it is a UXO, it can be recognized in the future, and if not, can be left in the ground.
The Metal Mapper is meant for use on land, but marine UXO detection and classification is becoming more important as offshore infrastructure, such as wind-farms, are built. In some settings magnetics is used, but it has the draw-back that it has no intrinsic depth information, and is incapable of detecting items that are not magnetic (such as many of the german-made aluminum bombs dropped in World Ware II). Building a TDEM system for offshore UXO detection is an application of interest for the future.
Wednesday was our first full day in Vienna, and getting to a yoga class was high on the priority list for the day! After going to the Geologic survey to see the room and test the setup for the DISC presentation, we found the Bikram studio in Vienna and attended a 4pm class. Sabine, the instructor, alternated between German and English (for the two Canadian tourists!). The studio was on the 4th floor and had a very pleasant atmosphere about it. Following yoga, we asked for a restaurant suggestion and ended up at a restaurant across the street, called Oben where we had a Thai chicken curry and a Korean noodle bowl with crispy pork — both were delicious!
Vienna is home to Mozart and Strauss. There are many venues throughout the city where you can attend a concert, and one is the Schoenbrunn Palace. Saturday morning, we looked at getting tickets for that evening. They were sold out except for the VIP section, so we decided to go for it (after all, there aren’t many opportunities in life to attend a concert in a palace, why not do so as a VIP!). The seats were fantastic — we were seated in the second row with a clear view of the conductor as he passionately swung his arms leading the orchestra.
We wanted to make the evening an Austrian experience and what could be more iconic than Wiener Schnitzel in Wien. We found a cozy restaurant that specializes in schnitzels. They serve a whole variety: some breaded, some not, some with sauces or filled with mushrooms. Between the two of us, we shared 2 different styles of Schnitzel, one classic and one with a tomato sauce. We managed to finish off our plates and certainly had no hunger pangs during the concert.
Austria closes on Sundays. Most of the grocery stores and local shops were closed; a few restaurants and shops in the city centre were open. We had a relaxed day and wandered through city, taking a few photos along the way. Sacher torte is an iconic treat in Austria. We found a cafe that is renowned for its Sacher torte and other treats, Cafe Central, and stopped in for an afternoon treat.
We managed to find a small grocery store with enough ingredients to make ourselves a pasta dinner that night, our last evening in Austria.
Thanks to Robert Supper for his support and efforts to gather a group of participants with diverse backgrounds!