Abstract. High alpine rock wall permafrost is extremely sensitive to climate change. Its degradation has a strong impact on landscape evolution and can trigger rockfalls constituting an increasing threat to socio-economical activities of highly frequented areas; quantitative understanding of permafrost evolution is crucial for such communities. This study investigates the long-term evolution of permafrost in three vertical cross sections of rock wall sites between 3160 and 4300 m above sea level in the Mont Blanc massif, from the Little Ice Age (LIA) steady-state conditions to 2100. Simulations are forced with air temperature time series, including two contrasted air temperature scenarios for the 21st century representing possible lower and upper boundaries of future climate change according to the most recent models and climate change scenarios. The 2-D finite element model accounts for heat conduction and latent heat transfers, and the outputs for the current period (2010)(2011)(2012)(2013)(2014)(2015) are evaluated against borehole temperature measurements and an electrical resistivity transect: permafrost conditions are remarkably well represented. Over the past two decades, permafrost has disappeared on faces with a southerly aspect up to 3300 m a.s.l. and possibly higher. Warm permafrost (i.e. > −2 • C) has extended up to 3300 and 3850 m a.s.l. in N and S-exposed faces respectively. During the 21st century, warm permafrost is likely to extend at least up to 4300 m a.s.l. on S-exposed rock walls and up to 3850 m a.s.l. depth on the N-exposed faces. In the most pessimistic case, permafrost will disappear on the S-exposed rock walls at a depth of up to 4300 m a.s.l., whereas warm permafrost will extend at a depth of the N faces up to 3850 m a.s.l., but possibly disappearing at such elevation under the influence of a close S face. The results are site specific and extrapolation to other sites is limited by the imbrication of local topographical and transient effects.
Modelling the discontinuity network of fractured reservoirs may be addressed (1) by purely stochastic means, (2) with a fractal approach, or (3) using mechanical parameters describing the spatial organisation of fracture systems. Our paper presents an approach where the geometrical properties of the fracture networks are incorporated in the form of both statistical and mechanical rules. This type of approach is particularly suitable to model stratified fractured rock masses comprising two orthogonal families of joints and a family of sedimentary discontinuities. Their geometrical arrangement is governed by two kinds of rules based on (1) statistical parameters such as the mean, standard deviation of joint length and of bed thickness, both determined by field observations, and (2) geometrical parameters that result from genetic processes inferred from field observations and analogue experiments on the nucleation and propagation mechanisms of joints. Using these parameters, we generate realistic networks in terms of the relative position of joints that control the overall network connectivity: the model enables all combinations of joint spacing and vertical persistence for orthogonal patterns ranging from ladder type to grid type patterns. It also integrates the concept of mechanical "saturation" of a bed, thereby permitting the generation of both "saturated" and "unsaturated" networks.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.