Abstract. Saharan dust deposits can turn snow covered mountains into a spectacular orange landscape. When avalanches release, a formerly buried dust layer can become apparent, possibly marking the base of the crown. This appearance may suggest a relation between avalanche release and the prior deposited dust, which found mention among recreationists and avalanche professionals alike. While dust deposition affects the absorption of solar energy altering snowpack temperatures and melt rates, to date, there is no clear scientific evidence that dust deposition can significantly modify avalanche activity. Here we investigate, using an ensemble snow cover model, the impact of dust deposition on snow properties and mechanical stability by comparing simulations with and without dust deposition for synthetic and observed dust deposition events. The study focuses on two typical avalanche situations: artificial triggering on persistent weak layers and natural release of wet-snow avalanches. We study several situations with and without dust deposition and demonstrate how sensitive the impact of dust deposition is to the deposited dust mass, the slope aspect, the elevation and the meteorological conditions following the dust deposition. The additional energy absorbed by the dust layer speeds up warming and may advance surface wetting to ease the formation of a melt-freeze crust. If the crust is buried, the phenomenon of a strong temperature gradient close to the crust may promote the formation of persistent weak layers inside the snowpack after weak layer burial. On the other hand, the melt-freeze crust may also lead to an increase of snowpack stability by redistributing the stress applied to buried weak layers. Regarding wet-snow instabilities, we show that dust deposition can advance the onset of wet-snow avalanche activity by up to one month in spring, as hypothesized in previous studies. Thus, the impact of Saharan dust deposition on snow mechanical stability can be either neutral, positive or negative, depending on the local snow and meteorological conditions. Even though not all physical processes are implemented, state-of the art snow cover models are able to mimic the speed-up of crust formation and snow instability models can point out relevant situations for avalanche forecasting.

Abstract. Saharan dust deposits can turn snow-covered mountains into a spectacular orange landscape. When avalanches release, a formerly buried dust layer can become apparent, possibly marking the failure plane. This appearance may suggest a relation between avalanche release and the previously deposited dust, which found mention among recreationists and avalanche professionals alike. While dust deposition affects the absorption of solar energy altering snowpack temperatures and melt rates, to date, there is no clear scientific evidence that dust deposition can significantly modify snow stability. Here we investigate, using an ensemble snow cover model, the impact of dust deposition on snow properties and mechanical stability by comparing simulations with and without dust deposition for synthetic and observed dust deposition events. The study focuses on two typical avalanche situations: artificial triggering on persistent weak layers and natural release of wet-snow avalanches. We study several situations with and without dust deposition and demonstrate how sensitive the impact of dust deposition is to the deposited dust mass, the slope aspect, the elevation and the meteorological conditions following the dust deposition. The additional energy absorbed by the dust layer speeds up warming and may advance surface wetting to ease the formation of a melt-freeze crust. If the crust is buried, the phenomenon of a strong temperature gradient close to the crust may promote the formation of persistent weak layers inside the snowpack. On the other hand, the melt-freeze crust may also lead to an increase in snowpack stability by redistributing the stress applied to weak layers buried below. Regarding wet-snow instability, we show that dust deposition can advance the onset of wet-snow avalanche activity by up to 1 month in spring, as hypothesized in previous studies. Thus, the impact of Saharan dust deposition on snowpack stability can be either neutral, positive or negative, depending on the topographical, snow and meteorological conditions. Even though not all physical processes are implemented, state-of the art snow cover models are able to mimic the speed-up of crust formation, and snow instability models can point out relevant situations for avalanche forecasting.

<p><span lang="en-US">The </span><span lang="en-US">thickness integrated</span><span lang="en-US"> dense flow avalanche simulation module com1DFA of the open source framework AvaFrame is used for snow avalanche simulations </span><span lang="en-US">with application in</span><span lang="en-US"> hazard mapping </span><span lang="en-US">for</span><span lang="en-US"> different mountainous areas. In order to further increase the information </span><span lang="en-US">value</span><span lang="en-US"> gained from the avalanche simulation results </span><span lang="en-US">in a global coordinate system</span><span lang="en-US">, we introduce a thalweg following coordinate system. It allows us to quantitatively compare simulation scenarios and results </span><span lang="en-US">of</span><span lang="en-US"> different</span><span lang="en-US"> modelling approaches </span><span lang="en-US">in a new way. </span><span lang="en-US">It helps to</span><span lang="en-US"> bridge the gap between the </span><span lang="en-US">modules</span><span lang="en-US"> operating in three-dimensional terrain </span><span lang="en-US">(com1DFA) </span><span lang="en-US">versus two-dimensional along the avalanche path, </span><span lang="en-US">such as the well-known alpha-beta model implemented in module com2AB</span><span lang="en-US">. </span><span lang="en-US">One </span><span lang="en-US">essential</span><span lang="en-US"> step of the analysis procedures (analysis modules in AvaFrame) is the avalanche </span><span lang="en-US">thalweg</span> <span lang="en-US">generation itself. The thalweg</span><span lang="en-US"> depends on the main flow direction, a property of the avalanche event which is </span><span lang="en-US">strongly</span><span lang="en-US"> influenced by the terrain the avalanche flow will encounter. So far, the main flow direction is usually derived from observations or avalanche simulations, and the thalweg is </span><span lang="en-US">generated</span><span lang="en-US"> manually. However, the reproducibility of this method raises an issue, and manually </span><span lang="en-US">identifying</span><span lang="en-US"> the avalanche thalweg for every slope is unnecessarily time-consuming. </span></p> <p><span lang="en-US">In this work, we use com1DFA simulations in three dimensional terrain. We automatically generate the two-dimensional avalanche thalweg by extracting the centre of mass coordinates at every time step. Projecting the simulation results into this thalweg following coordinate system, we can derive the position of the avalanche front and the local travel angles, from which scalar measures like runout length and runout angle are determined. We combine temporal and spatial information by introducing the thalweg-time and thalweg-altitude diagrams. These offer a different perspective on the simulation results and, at a glance, provide information on the evolution of spatio-temporal flow variables (thickness, velocity) along the avalanche thalweg in a single plot. Additionally, by using a numerical particle-grid method, we can evaluate simulation outputs at a particle level and relate them to the whole avalanche flow. Another advantage of the analysis tools operating in the thalweg coordinate system is the possibility to compare simulation results with field measurements. For example, we present in-flow particle sensors trajectories and corresponding velocities recorded during field experiments to evaluate com1DFA simulation results and thereby help to improve the dense flow module. For different avalanche simulations, we show how these analysis modules provide a new way to summarize the complex spatio-temporal flow variables evolution in three dimensional terrain in a more intuitive two dimensional illustration along the </span><span lang="en-US">automatically generated thalweg.</span></p>

<div><span>Mineral dust and black carbon are potent drivers of the snow cover evolution. After their deposition on the snow surface, they can impact snow albedo and thus the snowpack evolution including the timing of snow-melt. While BC deposition is rather constant along the winter season, mineral dust deposition is more sporadic in the French Alps, subject to large dust outbreak events coming from Sahara. The dust deposition drastically changes the snow color, its absorption of solar energy and, as a consequence, modifies the internal temperature of the snow layers and their metamorphism. While mountain practitioners often report higher avalanche activities after dust deposition events, there is, up to now, no clear evidence neither from observations nor modelling that dust deposition enhances avalanche activity. Here, we investigate, using ensemble detailed snowpack simulations, the impact of dust outbreak on snow metamorphism, snow stratigraphy and mechanical stability by comparing simulations with and without dust deposition under several meteorological conditions. The results show that the dust deposition can impact the spatial and temporal distribution of the unstable slopes. The effect of the deposition largely depends on the timing of dust deposition with respect to subsequent snowfalls. It also depends on the elevation, the aspect and the time since deposition event. By using multiphysics simulations, we were able to assess the robustness of our conclusions with respect to snowpack modelling errors.</span></div>