Geosciences Column: Landslide risk in a changing climate, and what that means for Europe’s roads

Geosciences Column: Landslide risk in a changing climate, and what that means for Europe’s roads

If your morning commute is already frustrating, get ready to buckle up. Our climate is changing, and that may increasingly affect some of central Europe’s major roads and railways, according to new research published in the EGU’s open access journal Natural Hazards and Earth System Sciences. The study found that, in the face of climate change, landslide-inducing rainfall events will increase in frequency over the century, putting central Europe’s transport infrastructure more at risk.  

How do landslides affect us?

Landslides that block off transportation corridors present many direct and indirect issues. Not only can these disruptions cause injuries and heavy delays, but in broader terms, they can negatively affect a region’s economic wellbeing.

One study for instance, published in Procedia Engineering in 2016, examined the economic impact of four landslides on Scotland’s road network and estimated that the direct cost of the hazards was between £400,000 and £1,700,000. Furthermore the study concluded that the consequential cost of the landslides was around £180,000 to £1,400,000.

Such landslides can have a societal impact on European communities as well, as disruptions to road and railway networks can impact access to daily goods, community services, and healthcare, the authors of the EGU study explain.

Modelling climate risk

To analyse climate patterns and how they might affect hazard risk in central Europe, the researchers first ran a set of global climate models, simulations that predict how the climate system will respond to different greenhouse gas emission scenarios. Specifically, the scientists ran climate projections based on the Intergovernmental Panel on Climate Change’s A1B socio-economic pathway, a scenario defined by rapid economic growth, technological advances, reduced cultural and economic inequality, a population peak by 2050, and a balanced reliance on different energy sources.

They then determined how often the conditions in their climate projections would trigger landslide events specifically in central Europe using a climate index that estimates landslide potential from the duration and intensity of rainfall events. The index, established by Fausto Guzzetti of National Research Council of Italy and his colleagues, suggests that landslide activity most likely occurs when a rainfall event satisfies the following three conditions: the event lasts more than three days, total downpour is more than 37.3 mm and at least one day of the rainfall period experiences more than 25.6 mm.

The researchers also incorporated into their models data on central Europe’s road infrastructure as well as the region’s geology, including topography, sensitivity to erosion, soil properties and land cover.

Overview of a particularly risk-prone region along the lowlands of Alsace and the Black Forest mountain range: (a) location of the region in central Europe and median of the increase in landslide-triggering climate events for (b) the near future and (c) the remote future.

The fate of Europe’s roadways

The results of the researchers’ models suggest that the number of landslide-triggering rainfall events will increase from now up until 2100. Their simulations also find while that these hazardous rainfall events slightly increase in frequency between 2021 and 2050, the number of these occurrences will be more significant between 2050 and 2100.  

While the flat, low-altitude areas of central Europe will only experience minor increases in landslide-inducing rainfall activity, regions with high elevation, like uplands and Alpine forests, are most at risk, their findings suggest.

The study found that many locations along the north side of the Alps in France, Germany, Austria and the Czech Republic may face up to seven additional landslide-triggering rainfall events as our climate changes. This includes the Vosges, the Black Forest, the Swabian Jura, the Bergisches Land, the Jura Mountains, the Northern Limestone Alps foothills, the Bohemian Forest, and the Austrian and Bavarian Alpine forestlands.

The researchers go on to explain that much of the Trans-European Transport Networks’ main corridors will be more exposed to landslide-inducing rainfall activity, especially the Rhine-Danube, the Scandinavian-Mediterranean, the Rhine-Alpine, the North Sea-Mediterranean, and the North Sea-Baltic corridors.

The scientists involved with the study hope that their findings will help European policy makers make informed plans and strategies when developing and maintaining the continents’ infrastructure.  

GeoTalk: How will large Icelandic eruptions affect us and our environment?

GeoTalk: How will large Icelandic eruptions affect us and our environment?

Geotalk is a regular feature highlighting early career researchers and their work. In this interview we speak to Anja Schmidt, an interdisciplinary researcher at the University of Cambridge who draws from atmospheric science, climate modelling, and volcanology to better understand the environmental impact of volcanic eruptions. She is also the winner of a 2018 Arne Richter Award for Outstanding Early Career Scientists. You can find her on twitter at @volcanofile. 

Thank you for talking to us today! Could you introduce yourself and tell us a little more about your career path so far?

I was born and raised in Leipzig, Germany. I started my career completing an apprenticeship as an IT system engineer with the engineering company Siemens. I then decided to combine my interests in geology and IT by studying geology and palaeontology (with minors in Computing/IT and Geophysics) at the University of Leipzig in Germany. As part of my degree programme, I also studied at the University of Leeds’ School of Earth and Environment as an exchange student. I liked studying there so much I ended up returning to Leeds for a PhD.

My PhD on the atmospheric and environmental impacts of tropospheric volcanic aerosol again combined my interests in computing and volcanology, although I had to educate myself in atmospheric physics and chemistry, which wasn’t easy to begin with. However, I was embedded in a diverse,   supportive research group with excellent supervision, which eased the transition from being a geologist to becoming a cross between an atmospheric scientist and a volcanologist.

Initially, being neither one nor the other made me nervous. My supervisors and mentors all had rather straightforward career paths, whereas I was thought of as an atmospheric scientist when I presented my research in front of volcanologists and as a volcanologist when I presented to atmospheric scientists.

After my PhD, I spent just under 2 years at one post-doc before securing an independent research fellowship at the University of Leeds. The first year of total independence and responsibility as principle investigator was very challenging, but after a while I began to appreciate the benefits of the situation. I also really started to embrace the fact that I would always sit between the disciplines. I spent my summers in the United States at the National Centre for Atmospheric Research, helping them to build up their capability to simulate volcanic eruptions in their climate model. These research visits had a major impact on my career as they generated a lot of new research ideas, opened up opportunities and strengthened my network of collaborators greatly.

I considered myself settled when, shortly before the end of my fellowship, a lectureship came up. It had the word ‘interdisciplinary’ in its title and I simply couldn’t resist. Since September 2017, I have been an interdisciplinary lecturer at the University of Cambridge in the UK.

At this year’s General Assembly, you will receive an Arne Richter Award for Outstanding Early Career Scientists for your work on the environmental impacts of volcanic eruptions. What brought you to study this particular field?

I have always been fascinated by volcanic eruptions, but my first active volcano viewing wasn’t until college, where I had to chance to travel to Stromboli, a volcanic island off the coast of Sicily. While studying at the University of Leipzig, I used every opportunity to join field trips to volcanoes. I ended up spending 10 weeks in Naples, Italy to work with Giovanni Chiodini, a researcher from the National Institute of Geophysics and Volcanology in Rome, and his team on CO2 degassing from soils at the Solfatara volcano. Later on I was awarded a scholarship from the University of Leeds, which allowed me to delve deeper into the subject, although I ended up learning as much about atmospheric science and computer modelling as about volcanology.

Anja in front of the 2010 Fimmvörðuháls eruption in Iceland. Fimmvörðuháls was the pre-cursor eruption to Eyjafjallajökull. Credit: Anja Schmidt.

My PhD work focused on Icelandic volcanism and its potential effects on the atmosphere as well as society. In 2010, during the 3rd year of my PhD studies, Eyjafjallajökull erupted in Iceland. While an eruption like this and its impacts did not really come as a surprise to a volcanologist, I personally considered it a game-changer for my career. I had an opportunity to witness the pre-cursor eruption in Iceland and present my research. Within a matter of months, interest in my work increased. I even started to advise UK government officials on the risks and hazards of volcanic eruptions in Iceland.

In August 2014, an effusive eruption started at the Holuhraun lava field in Iceland. To this date, analysing field measurements and satellite data of the site and modelling simulations keeps me busy. Many of my senior colleagues told me that there is one event or eruption that defined their careers; for me that’s the 2014-2015 Holuhraun eruption.

At the General Assembly you also plan to talk about your work on volcanic sulphur emissions and how these emissions can alter our atmosphere as well as potentially affect human health in Europe. Could you tell us a little more about this research?

On average, there is one volcanic eruption every three to five years in Iceland. The geological record in Iceland also reveals that sulphur-rich and long-lasting volcanic eruptions, similar to Iceland’s Laki eruption in 1783-1784, occur once every 200 to 500 years. Sulphur dioxide and sulphate particles produced by volcanic eruptions can have detrimental effects on air quality and human health. Historical records from the 1780s imply that the Laki eruption caused severe environmental stress and contributed to spikes in mortality rates far beyond the shores of Iceland. While these long-lasting eruptions occur much less frequently than more typical short-duration explosive eruptions (like Grímsvötn 2011), they are classified as ‘high-impact’ events.

I was always interested in investigating how a similar magnitude eruption like Laki’s would affect modern society. By combining a global aerosol microphysics model with volcanological datasets and epidemiological evidence, I led a cross-disciplinary study to quantify the impact that a future Laki-type eruption would have on air quality and human health in Europe today.

Our work suggests that such an eruption could significantly degrade air quality over Europe for up to 12 months, effectively doubling the concentrations of small-sized airborne particles in the atmosphere during the first three months of the eruption. Drawing from the epidemiological literature on human response to air pollution, I showed that up to 140,000 cardiopulmonary fatalities could occur across Europe due to such an eruption, a figure that exceeds the annual mortality from seasonal influenza in Europe.

In January 2012, this discovery was used by the UK government as contributing evidence for including large-magnitude effusive Icelandic eruptions to the UK National Risk Register. This will help to mitigate the societal impacts of future eruptions through contingency planning.

Anja and her colleague Evgenia Ilyinskaya from the University of Leeds carrying out measurements during the 2014-2015 Holuhraun eruption in Iceland. Credit: Njáll Fannar Reynisson.

Since then, we have done more work on smaller-magnitude effusive eruptions such as the 2014-2015 Holuhraun eruption in Iceland, showing that this eruption resulted in short-lived volcanic air pollution episodes across central and northern Europe and longer-lasting and more complex pollution episodes in Iceland itself.

Something that you’ve touched on throughout this interview are the challenges of ‘sitting between the disciplines.’ From your experience, what has helped you address these issues throughout your career?

Indeed, it is often challenging to sit between the disciplines, but it can also be very rewarding. It helps to ignore boundaries between disciplines. I also tend to read a lot and very widely to get an idea of key concepts and issues in specific fields. In addition, I think collaboration and a willingness to challenge yourself are key if you want to make progress and break traditional disciplinary boundaries.

Anja, thank you so much for speaking to us about your research and career path. Before I let you go, what advice do you have for aspiring scientists? 

Be curious and never hesitate to ask a lot questions, no matter how ‘stupid’ or basic they may seem to you. The latter is particularly true when it comes to cross-disciplinary collaboration and work.  I also didn’t always follow the conventional route most people would advise you to take to achieve something. Never be afraid to take a chance or work with some level of risk.

I also have two or three close mentors that I can approach whenever I require some advice or feedback. No matter what career stage you are at, I think it almost always helps to get an outsider’s perspective and insight not only when there are problems.

Finally, never forget to have fun. Some of my best pieces of work were done when I was surrounded by collaborators that are really fun to be with and work with!

Interview by Olivia Trani, EGU Communications Officer.


Ilyinskaya, E., et al.: Understanding the environmental impacts of large fissure eruptions: Aerosol and gas emissions from the 2014–2015 Holuhraun eruption (Iceland), Earth and Planetary Science Letters, 472, 309-322, 2017

Schmidt, A., et al.: Satellite detection, long-range transport, and air quality impacts of volcanic sulfur dioxide from the 2014–2015 flood lava eruption at Bárðarbunga (Iceland)Journal of Geophysical Research: Atmospheres12097399757, 2015

Schmidt, et al.: Excess mortality in Europe following a future Laki-style Icelandic eruption, Proceedings of the National Academy of Sciences, 108(38), 15710-15715, 2011

Geosciences Column: The science behind snow farming

Geosciences Column: The science behind snow farming

For roughly the last decade, some ski resorts and other winter sport facilities have been using a pretty unusual method to ensure white slopes in winter. It’s called snow farming. The practice involves collecting natural or artificially made snow towards the end of winter, then storing the frozen mass in bulk over the summer under a thick layer of sawdust, woodchips, mulch, or other insulating material.

Many winter sport destinations have adopted the practice. In preparation for the 2014 Winter Olympics, Sochi, Russia stockpiled about 800,000 cubic metres of human-made snow during the warmer season, enough snow to fill 320 Olympic-size swimming pools.

Despite the growing trend, there still is little research on snow farming techniques. Recently, a team of scientists from the Institute for Snow and Avalanche Research (SLF) and the CRYOS Laboratory at the École Polytechnique Fédérale in Switzerland examined the success of snow conservation practices and used models to estimate what factors influence covered snow. Their findings were published in the EGU’s open access journal The Cryosphere.

Why store snow for the winter?

The ski industry has been storing snow for many reasons. The practice is a way for winter sports facilities to accommodate training athletes, start ski seasons earlier, and guarantee snow for major sports events. Snow farming can also be seen as a way to adapt to Earth’s changing climate, according to the authors of the study. Indeed, research published last year in The Cryosphere, found that the Alps may lose as much as 70 percent of snow cover by the end of the century if global warming continues unchecked. Snow loss to this degree could severely threaten the $70 billion dollar (57 billion EUR) industry and the alpine communities that depend on ski tourism.

For some ski resorts, the effects of climate change are already visible. For example, in Davos, Switzerland, a popular venue of the International Ski Federation Cross-Country World Cup, winter temperatures have risen over the last century while snow depth in turn has steadily declined.

Snow heap study

The research team studied two snow heaps: one near Davos, Switzerland (pictured here) and another in South Tyrol. Credit: Grünewald et al.

To better understand snow conservation techniques, the research team studied two artificially made snow heaps: one sitting near Davos and another located in South Tyrol. Each pile contained approximately 7,000 cubic metres of snow, about enough ice and powder to build 13,000 1.8-metre tall snowmen. The piles were also each covered with a 40 cm thick layer of sawdust and chipped wood.

Throughout the 2015 spring and summer season, the researchers measured changes in snow volume and density, as well as recorded the two sites’ meteorological data, including air temperature, humidity, wind speed and wind direction. The research team also fed this data to SNOWPACK, a model that simulates snow pile evolution and helps determine what environmental processes likely impacted the snow.

Cool under heat

From their observations, the researchers found that the sawdust and chipped wood layering conserved more than 75 percent of the Davos snow volume and about two thirds of the snow in South Tyrol. Given the high proportion of remaining snow, the researchers conclude that snow farming appears to be an effective tool for preparing for winter.

According to the SNOWPACK model, while sunlight was the biggest source of snow melt, most of this solar radiation was absorbed by the layer of sawdust and wood chips. The simulations suggest that the snow’s covering layer took in the sun’s heat during the day, then released this energy at night, creating a cooling effect on the snow underneath. Even more, the model found that, when the thick layer was moist, the evaporating water cooled the snow as well. The researchers estimate that only nine percent of the sun’s energy melted the snow heaps. Without the insulating layer, the snow would have melted far more rapidly, receiving 12 times as much energy from the sun if uncovered, according to the study.

Images of the South Tyrol snow heap from (a) 19 May and (b) 28 October. The snow depth (HS) is featured in c & d and snow height change (dHS) is shown in e. Credit: Grünewald et al.

The researchers found that the thickness of the covering layer was an important factor for snow conservation. When the team modelled potential snow melt under a 20 cm thick cover, the insulating and cooling effects from the layer had greatly diminished.

The simulations also revealed that, while higher air temperatures and wind speed increased snow melt, this effect was not very significant, suggesting that subalpine areas could also benefit from snow farming practices.

In the face of changing climates and disappearing snow, snow farming may be one solution for keeping winters white and skiers happy.


Grünewald, T., Wolfsperger, F., and Lehning, M.: Snow farming: conserving snow over the summer season, The Cryosphere, 12, 385-400,, 2018.

Marty, C., Schlögl, S., Bavay, M., and Lehning, M.: How much can we save? Impact of different emission scenarios on future snow cover in the Alps, The Cryosphere, 11, 517-529,, 2017.



Record-setting forest fires in 2017 – what is to blame?

Record-setting forest fires in 2017 – what is to blame?

Forest fires have once again seized the public consciousness in both Europe and North America. Extreme drought and temperatures contributed to a tinderbox in many forests, and have led to deadly fires across Europe and record-breaking, highly disruptive fires in the USA and Canada, from where I’m currently writing.

A simple way to understand fire is by thinking about the fire triangle – the three pieces that need to combine to produce fire: heat, oxygen, and fuel. While the concept of the fire triangle offers simple insights into the cause of a fire, each of the three factors are subject to a number of controls.

The availability of fuel in particular is affected by a whole range of influences, from tree species variations, to the impact of pests, to the prior history of wildfires. In many cases there will be a multitude of reasons for both the rate that a given fire spreads, and the spatial extent to which it burns. The extensive media coverage of this year’s record-breaking fire season provides a useful opportunity to explore some of these factors, and how they might have contributed to the fires that we’re seeing this year.

Multi-tiered fires

Not all forest fires are equal. The forestry service in British Columbia defines 6 different levels of fire, ranging from slow burning of peat below the surface of a forest (a ground fire), through the burning of scrub and ground level debris (surface fire), to far harder to control conflagrations that consume the canopy of the trees (a crown fire).

Each type of fire consumes fuel from different levels of the forest, and a transition from a lower to higher level requires at a minimum that there is enough flammable material in the upper layers of the forest. As such, the history of fire in a given stand of woodland will have a significant effect on the potential for future fires; a prior surface fire could remove the under-brush and limit the future fire risk.

Many older trees in a forest can have fire scars from previous, smaller fires that have not burnt them entirely, demonstrating that not all fires make the transition to crown fires. In the US, the largest 2-3% of wildfires contribute 95% of the total area burned annually, and these are generally the largest crown fires. For these large fires, the conditions must be optimal to reach the canopy, but once they make this transition they can be very difficult to stop.

Although the largest fires require a specific set of conditions, the transition to large fire can lag for a long time behind the initial trigger of a fire:

“Once ignited, decaying logs are capable of smouldering for weeks, or even months, waiting the time when prevailing conditions (hot, windy, and dry) are conducive for expansion into a full-blown forest fire,” write Logan & Powell in 2001.

The initial trigger in natural settings for the fire is generally lightning, or in rare cases the intense heat from lava flows. However, a recent study has shown that in the continental US, 84% of fires are started by humans. This can range from discarded cigarettes to prescribed fires that have raged out of control.

So, we now have a simplified understanding of the requirements for a fire, but perhaps the more important question remains: why do they spread?

What fans the flames?

So what are the processes that control the strength and spread of a fire? Certain aspects, like climatic conditions, tend to set a long-term propensity for wildfire, while other short term effects define the local, immediate cause for a fire.

The long-term role of climate is quite diverse. Flannigan and co-authors, in their 2008 paper, summarise the importance of temperature in setting the conditions for fire:

“First, warmer temperatures will increase evapotranspiration, as the ability for the atmosphere to hold moisture increases rapidly with higher temperatures… (Roulet et al. 1992) and decreasing fuel moisture unless there are significant increases in precipitation. Second, warmer temperatures translate into more lightning activity that generally leads to increased ignitions (Price and Rind 1994). Third, warmer temperatures maylead to a lengthening of the fire season (Westerling et al. 2006).”

The importance of temperature at a global scale is demonstrated by studies suggesting that fire outbreaks have systematically increased since the last glacial maximum 21,000 years ago, when average temperatures were several degrees colder.

So once the climatic conditions set a long term likelihood for a fire to break out, as well as the distribution of tree species (and therefore the amount of fuel), short term disturbances can act to further increase susceptibility. We can divide these more immediate factors into weather, ecological, and human influences.

Unsurprisingly, weather conditions that bring hot, dry weather with the chance of thunderstorms will be highly likely to drive fires. A number of studies have shown that this kind of weather is often linked to persistent high pressure systems in the atmosphere, which push away rainfall, and can last for weeks at a time.

This kind of atmospheric conditions can often be linked to longer term weather patterns, in particular El Niño, and the effects can be long-lived. For example, El Niño brings warm, dry weather to the North-western part of North America, driving increased fire. In the South-west, El Niño brings wetter weather, but the increased vegetation growth provides a larger amount of fuel that can then burn in subsequent drier conditions.

Caption: Healthy pines mix with red, decaying trees afflicted with pine beetle infestation in Jasper National Park, Canada (Aug 2017). Credit: Robert Emberson

The growth of plants is just one way that the forest ecosystem can affect the availability of fuel for fire. Many trees are affected by pests. The classic example is the pine beetle, which can kill so many trees that whole swathes of pine forest are left characteristically red as they decay in the aftermath of an infestation. While some studies have shown that this can reduce the chance of a crown fire (since less fuel is available in the canopy), the dead and decaying trees provide a source of drier fuel at ground level that is a concern in many regions.

Elsewhere, we can see cases where trees encourage fire. Eucalyptus trees contain oil that burns strongly; the fire that this produces is suggested to remove the other tree species competing with the Eucalyptus. A strange method, but no doubt effective! The Eucalyptus, introduced to Portugal by humans in the 18th century, has been linked to many of the deadly fires that occurred there this summer. The lodgepole pine, too, has pinecones that require the heat of fire to open and release their seeds.
As such, fire is a natural part of the ecosystem, and in most areas of the boreal forest fire management is limited, with attempts at suppression only made where human settlement is at risk.

Where humans do step in, their actions can have an important role in setting the overall susceptibility to fire. Creating ‘fire breaks’ by felling or controlled burning of woodland in the path of a fire removes fuel and limits the growth of a fire, and a program of controlled burns is an important part of forest management to limit the potential for future fire by clearing scrub vegetation at the surface.

On the other hand, continual suppression of fires can lead to build up of fuel – which can then form a significantly more dangerous fire on a long term basis.

Ken Lertzman, Professor of Forest Ecology and Management at Simon Fraser University, told me that in general, control mechanisms are only useful for smaller surface fires; once the fire reaches the canopy, fire suppression is extremely difficult.

“It’s a combination of a statistical and philosophical problem”, he says. Fire control is expensive, so unless it’s financed by profits from felled lumber, cost benefit analysis is necessary, based in part on the statistical probability of a huge fire breaking out in a given location. Philosophically, though, “management forest stand structure at the boundary conditions [between surface and canopy fires] may be able to keep small fires becoming extreme.” and thus perhaps it’s worth trying to use control mechanisms even if giant canopy fires would ignore them, just to avoid that transition to the canopy.

As with most natural systems, the factors discussed above don’t necessarily act independently. Many can amplify other effects. For example, the geographical range of pine beetle outbreaks could increase under a warming climate. At a smaller scale, giant fires can create their own weather patterns, acting to dry out the surrounding forest even before it ignites. So, is the greater incidence of fire this year down to this kind of combination of factors?

What’s happening in 2017?

The sun at 10.15 in the morning in Chilliwack, BC, 3rd August 2017 – more than 50km from the nearest fire. Credit: Megan Reich.

The summer of 2017 has been brutal for wildfire in many locations. Europe has been hit hard with more fires in one day in Portugal than any previous on record. In British Columbia (BC), the largest single fire on record is currently burning, contributing to the largest total area burned in historical record. The fires released so much particulate matter and haze that the sun was obscured in large parts of BC.

According to Professor Lertzman, in British Columbia at least, this is “a fire year different in degree, not in kind”. The same processes are at play as always, but in overdrive. An early spring thaw combined with a long, hot and dry summer created the ideal antecedent conditions for fire.

Linking individual intense ‘fire years’ to long term climate change in challenging, but it is likely that these kind of conditions would be more likely to occur in a world influenced by anthropogenic climate change. Years like this one, which are clearly exceptional compared to the long term trend for fire, might begin to occur more often; every 5-10 years rather than every 30-50, for example.

In the short term, fire damage and suppression is expensive, both financially and in terms of lives affected. Smoke is a health risk to wide areas in the face of such intense fire, but ecological damage can be more difficult for humans to see.

Mature forest is a different niche for organisms to fill in comparison to fresh surfaces stripped bare by fire. While each ecosystem has its place in the natural forest, an increased prevalence of fire reduces the mature forest available for the species that prefer that ecological zone. These can range from recognisable mammal species such as bears, deer and caribou, to less well-known but still important canopy lichen species, says Professor Lertzman.

While this year’s fire season is beginning to ease off, it is clear that the range of factors, both natural and those driven by humans, will continue to play a role in years to come. More build-up of infrastructure in developed countries puts more human settlement at risk, so a clear understanding of how fire interacts with the climate, weather, and forest management strategies will be vital to allow us to live alongside fire in the future with fewer problems.

By Robert Emberson, freelance science writer


Further reading and resources