One fifth of Earth's volcanoes are covered by snow or ice and many have active geothermal systems that interact with the overlying ice. These glaciovolcanic interactions can melt voids into glaciers, and are subject to controls exerted by ice dynamics and geothermal heat output. Glaciovolcanic voids have been observed to form prior to volcanic eruptions, which raised concerns when such features were discovered within Job Glacier on Qw̓elqw̓elústen (Mount Meager Volcanic Complex), British Columbia, Canada. In this study we model the formation, evolution, and steady-state morphology of glaciovolcanic voids using analytical and numerical models. Analytical steady-state void geometries show cave height limited to one quarter of the ice thickness, while numerical model results suggest the void height h scales with ice thickness H and geothermal heat flux $\dot {Q}$ as $h/H = a H^b \dot {Q}^c$, with exponents b = −n/2 and c = 1/2 where n is the creep exponent. Applying this scaling to the glaciovolcanic voids within Job Glacier suggests the potential for total geothermal heat flux in excess of 10 MW. Our results show that relative changes in ice thickness are more influential in glaciovolcanic void formation and evolution than relative changes in geothermal heat flux.

This work proposes a holistic quantitative snow avalanche risk assessment that evaluates, at reasonable computational costs and for various types of buildings, the impact of forest cover changes on the probability distribution of runout distances, impact pressures and subsequent risk estimates. A typical case study of the French Alps shows that, from a completely deforested to a completely forested path, avalanche risk for a building located downslope decreases by 53–99%, depending on how forest cover is accounted for in avalanche statistical–dynamical modeling. Local forest cover data inferred from old maps and photographs further demonstrates that a 20–60% risk reduction actually occurred between 1825 and 2017 at the site because of the afforestation dynamics, with significant modulations according to the considered building technology. These results (1) assert the protective role of forests against snow avalanches, (2) highlight the potential of combining nature-based solutions with traditional structural measures to reduce risk to acceptable levels at reasonable costs, (3) suggest a significant decrease in risk to settlements in areas that encountered similar forest cover changes and (4) open the door to the quantification of long-term avalanche risk changes as a function of changes of all its hazard, vulnerability and exposure drivers in various mountain context.

Physically-based avalanche propagation models must still be locally calibrated to provide robust predictions, e.g. in long-term forecasting and subsequent risk assessment. Friction parameters cannot be measured directly and need to be estimated from observations. Rich and diverse data are now increasingly available from test-sites, but for measurements made along flow propagation, potential autocorrelation should be explicitly accounted for. To this aim, this work proposes a comprehensive Bayesian calibration and statistical model selection framework. As a proof of concept, the framework was applied to an avalanche sliding block model with the standard Voellmy friction law and high rate photogrammetric images. An avalanche released at the Lautaret test-site and a synthetic data set based on the avalanche are used to test the approach and to illustrate its benefits. Results demonstrate (1) the efficiency of the proposed calibration scheme, and (2) that including autocorrelation in the statistical modelling definitely improves the accuracy of both parameter estimation and velocity predictions. Our approach could be extended without loss of generality to the calibration of any avalanche dynamics model from any type of measurement stemming from the same avalanche flow.

Glacier thinning and retreat drives initial acceleration of glacier sliding and erosion, de-buttressing of steep valley walls, and destabilization of ice-marginal deposits and bedrock, which can lead to massive rock avalanching and accelerated incision of tributary watersheds. A compelling example of these changes occurred in Taan Fjord in SE Alaska due to the rapid thinning and retreat of Tyndall Glacier over the past half century. Increased glacier sliding speeds led to both increased rates of subglacial erosion and the evacuation of subglacially stored sediments into the proglacial basins. The shrinking glacier also exposed proglacial tributary watersheds to rapid incision and denudation driven by >350 m of baselevel fall in a few decades. Moreover, in October 2015 a large tsunamigenic landslide occurred at the terminus of Tyndall Glacier, largely due to thinning exposing oversteepened, unstable slopes. Sediment yields from the glacier, the landslide and the tributary watersheds, measured from surveys of the sediments in the fjord collected in 1999 and 2016, are compared to ongoing changes in glacier and fjord geometry to investigate the magnitude of glacial and paraglacial denudation in Taan Fjord during retreat. In the last 50 years, sediment yields from the glacier and non-glacial tributaries kept pace with the rapid rate of retreat, and were on par with each other. Notably, basin-averaged erosion rates from the paraglacial landscape were twice that from the glacier, averaging 58 ± 9 and 26 ± 5 mm a−1, respectively. The sharp increases in sediment yields during retreat observed from both the glacier and the adjacent watersheds, including the landslide, highlight the rapid evolution of landscapes undergoing glacier shrinkage.

Very large volumes of ice break off regularly from Taconnaz hanging glacier, French Alps. During winter, when the snow mantle is unstable, these collapses can trigger very large avalanches that represent a serious threat to inhabited areas below. Photogrammetric measurements have been performed over 1 year to assess the volume and frequency of the largest collapses. Major collapses occur when the glacier reaches a critical geometry. After a major ice collapse, the glacier is in a minimal position and subsequently recharges over 6 months to reach the maximum position again. This critical geometry is a necessary but not sufficient condition for further large collapses. Large collapses do not systematically occur in the maximum position, as ice is often removed by disintegration into small ice blocks. For two major collapses, the volume of ice breaking off has been assessed at ~275 000 m3. Photogrammetric measurements were used to determine an ice flux of 820 000 m3 a–1 through the studied ice stream, in agreement with an assessment based on ice-flow modeling. This ice flux estimation was used to determine the average ice volumes breaking off over surveyed periods.

The deadliest outburst flood from an englacial cavity occurred on Glacier de Tête Rousse in the Mont Blanc area, French Alps, in 1892. A subglacial reservoir was discovered in the same glacier in 2010 and drained artificially in 2010, 2011 and 2012 to protect the 3000 inhabitants downstream. The mechanism leading to the spontaneous refilling of the cavity following these pumping operations has been analyzed. For this purpose, the subglacial water volume changes between 2010 and 2013 were reconstructed. The size of the cavity following the pumping was found to have decreased from 53500 m3 in 2010 to 12 750 m3 in 2013. Creep and the partial collapse of the cavity roof explain a large part of the volume loss. Analysis of cavity filling showed a strong relationship between measured surface melting and the filling rate, with a time delay of 4–6 hours. A permanent input of 15 m3 d−1, not depending on surface melt, was also found. The meltwater and rain from the surface is conveyed to bedrock through crevasses and probably through a permeable layer of rock debris at the glacier bed. The drainage pathway permeability was estimated at 0.054 ms−1 from water discharge measurements and dye-tracing experiments.

To better understand the risk of disasters due to glacial lake outburst floods (GLOFs), we synthetically analyze the spatial distribution and evolution of moraine-dammed lakes and potentially dangerous glacial lakes (PDGLs) in the Chinese Himalaya. Our county-based assessment of GLOF disaster risk combines PDGL outburst hazard, regional exposure, vulnerability of exposed elements and adaptation capability (risk management) using the analytic hierarchy process. We synthetically analyze the disaster risk using the weighted comprehensive method. Remote-sensing data show there are 329 moraine-dammed lakes (>0.02 km2; total area 125.43 km2) in the Chinese Himalaya, of which 116 (total area 49.49 km2) are identified as PDGLs. The zones at highest risk of GLOF disaster are mainly located in Nyalam, Tingri, Dinggyê, Lhozhag, Kangmar and Zhongba, in the mid-eastern Himalaya. Lowest-risk zones are located in the eastern Himalaya. On the county scale, Lhozhag and Lhunze have the highest hazard degrees and exposure, while Zhongba and Zando have the highest degree of vulnerability and lowest adaptation capacity. Our regionalization results for GLOF disaster risk are consistent with the distribution of historical disaster sites across the Chinese Himalaya.