Snow avalanche‐induced air‐blasts are capable of breaking trees, damaging buildings and causing fatalities. Predicting their destructive properties is an essential part of snow avalanche hazard mitigation. Here, we propose a depth‐averaged model that involves turbulent fluctuations to simulate the air‐blast dynamics. The turbulent energy of the air‐blast arises from that of dust‐mixed air transferred from the avalanche core, shearing work in the cloud and entrained air, and is exploited to improve the air entrainment and drag relationships. We further present a unique data set of air blast‐induced tree breakage, providing type, status, diameter and falling direction of the measured trees. Through case studies of two artificially released avalanches with measured powder heights and three natural avalanches with tree‐breakage information, we test the model and investigate the turbulence effect on air‐blast dynamics. The proposed model and tree‐breakage data set quantify the air‐blast destructiveness and can be applied for avalanche hazard assessment.
23 October 2024 in Natural Hazards and Earth System Sciences
Natural disturbances like windthrows or forest fires alter the provision of forest ecosystem services such as timber production, protection from natural hazards, and carbon sequestration. After a disturbance, forests release large amounts of carbon and therefore change their status from carbon sinks to carbon sources for some time. Climate-smart forest management may decrease forest vulnerability to disturbances and thus reduce carbon emissions as a consequence of future disturbances. But how can we prioritise the stands most in need of climate-smart management? In this study we adopted a risk mapping framework (hazard times vulnerability) to assess the risk to climate-related forest ecosystem services (carbon stock and sink) in forests prone to windthrow (in the Julian Alps, Italy) and forest fires (in the Apennines, Italy). We calculated hazard by using forest fire and windthrow simulation tools and examined the most important drivers of the respective hazards. We then assessed vulnerability by calculating current carbon stocks and sinks in each forest stand. We combined these values together with the calculated hazard to estimate “carbon risk” and prioritised high-risk stands for climate-smart management. Our findings demonstrate that combining disturbance simulation tools and forest carbon measurements may aid in risk-related decision-making in forests and in planning decisions for climate-smart forestry. This approach may be replicated in other mountain forests to enhance our understanding of their actual carbon vulnerability to forest disturbances.
3 August 2020 in ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Increasingly advanced and affordable close-range sensing techniques are employed by an ever-broadening range of users, with varying competence and experience. In this context a method was tested that uses photogrammetry and classification by machine learning to divide a point cloud into different surface type classes. The study site is a peat scarp 20 metres long in the actively eroding river bank of the Rotmoos valley near Obergurgl, Austria. Imagery from near-infra red (NIR) and conventional (RGB) sensors, georeferenced with coordinates of targets surveyed with a total station, was used to create a point cloud using structure from motion and dense image matching. NIR and RGB information were merged into a single point cloud and 18 geometric features were extracted using three different radii (0.02 m, 0.05 m and 0.1 m) totalling 58 variables on which to apply the machine learning classification. Segments representing six classes, dry grass, green grass, peat, rock, snow and target, were extracted from the point cloud and split into a training set and a testing set. A Random Forest machine learning model was trained using machine learning packages in the R-CRAN environment. The overall classification accuracy and Kappa Index were 98% and 97% respectively. Rock, snow and target classes had the highest producer and user accuracies. Dry and green grass had the highest omission (1.9% and 5.6% respectively) and commission errors (3.3% and 3.4% respectively). Analysis of feature importance revealed that the spectral descriptors (NIR, R, G, B) were by far the most important determinants followed by verticality at 0.1 m radius.