Search

Large collections of images have become readily available through modern digital catalogs, from sources as diverse as historical photographs, aerial surveys, or user-contributed pictures. Exploiting the quantitative information present in such wide-ranging collections can greatly benefit studies that follow the evolution of landscape features over decades, such as measuring areas of glaciers to study their shrinking under climate change. However, many available images were taken with low-quality lenses and unknown camera parameters. Useful quantitative data may still be extracted, but it becomes important to both account for imperfect optics, and estimate the uncertainty of the derived quantities. In this paper, we present a method to address both these goals, and apply it to the estimation of the area of a landscape feature traced as a polygon on the image of interest. The technique is based on a Bayesian formulation of the camera calibration problem. First, the probability density function (PDF) of the unknown camera parameters is determined for the image, based on matches between 2D (image) and 3D (world) points together with any available prior information. In a second step, the posterior distribution of the feature area of interest is derived from the PDF of camera parameters. In this step, we also model systematic errors arising in the polygon tracing process, as well as uncertainties in the digital elevation model. The resulting area PDF therefore accounts for most sources of uncertainty. We present validation experiments, and show that the model produces accurate and consistent results. We also demonstrate that in some cases, accounting for optical lens distortions is crucial for accurate area determination with consumer-grade lenses. The technique can be applied to many other types of quantitative features to be extracted from photographs when careful error estimation is important.

International audience; Since the early 2000s, a remarkable amount of rockfalls has been observed in permafrost areas of the mid-latitude mountain ranges concurrently to hot summers. This study explores the seasonal thaw (ST) in permafrost rock walls of the Aiguille du Midi site (3842 m a.s.l., Mont Blanc massif). We first analyse six years of temperature records in three 10 m-deep boreholes against air temperature (AT) and a four-year time series of pictures showing the snow conditions on two rock faces. Then, we test the sensitivity of the active layer against eight snow fall scenarios using the 1-D surface energy balance and heat conduction model CryoGrid 3 forced by in-situ measurements from a vertical face. Snow falls occur all the year round at this elevation and play an important role for the active layer thickness (ALT), but the snow cover and its control are highly heterogeneous. A long-lasting of a snow cover during spring/early summer delays the ST and reduces the ALT. The thicker and the more spatially-continuous is the snow cover, the stronger are the delay and ALT reduction. Convective clouds could also reinforce this pattern. The summer AT and heat waves are the dominant controlling factors of the ALT. But summer snow falls can sometimes persist for several days on the rock surface and reduce the effect of the heat waves. Active layer can thicken during the early fall, except if the snow starts to accumulate on the rock surface and favours the refreezing. The timing of the snow fall is the most critical parameter to determine the snow effect on the ALT. This study suggests that the characteristics of the bedrock and snow accumulation (steepness, surface roughness, and sun-exposure) must be taken into account to better understand the formation and changes of the active layer and its possible implications for rockfall triggering.

International audience; The development of shallow landslides is strongly connected to the changes in the water content of soils on hillslopes, their clay content and permeability distribution, which, in turn, are playing an important role regarding their hydro-mechanical properties. A non-intrusive geophysical method able to map these properties would be very helpful. The most common geoelectrical method, DC (Direct Current) resistivity, cannot be used as a stand-alone technique for this purpose since it depends on two contributions (bulk and surface conductivities), which depend on the water content and the cation exchange capacity (CEC) of the material. Induced polarization is a geophysical method that can be now used to complement DC resistivity in providing key material properties that can be used to diagnose potential risks for failure. We first recall the basic principles behind the induced polarization method from laboratory to field scales and key findings in the underlying petrophysics needed to jointly interpret electrical conductivity and normalized chargeability tomograms. Then, we apply these relationships to a field survey carried out over a shallow landslide at Claix (Isère, France), close to Grenoble. A 3D induced polarization survey was carried out and interpreted in terms of the clay content, water content, and permeability distributions. We demonstrate that the landslide is associated with a channel of high water content corresponding with the presence of travertine, a flow-path, and a permeability barrier downslope corresponding to the presence of plastic clays. This study demonstrates that induced polarization can be used to characterize the impacted volume and therefore might have been useful to map the area before the landslide to assess the possible risk of failure. This methodology could play a key role in mitigation planning.