At the Junction Between Deep Learning and Statistics of Extremes: Formalizing the Landslide Hazard Definition

• Published in: Journal of geophysical research. Machine learning and computation Vol. 1; no. 3
• Main Authors: Dahal, Ashok, Huser, Raphaël, Lombardo, Luigi
• Format: Journal Article
• Language: English
• Published: 01.09.2024
• Subjects: climate change; deep learning; landslides; modeling; Nepal; statistics of extremes
• DOI: 10.1029/2024JH000164
• ISSN: 2993-5210

The most adopted definition of landslide hazard combines spatial information about landslide location (susceptibility), threat (intensity), and frequency (return period). Only the first two elements are usually considered and estimated when working over vast areas. Even then, separate models constitute the standard, with frequency being rarely investigated. Frequency and intensity are dependent on each other because larger events occur less frequently and vice versa. However, due to the lack of multi‐temporal inventories and joint statistical models, modeling such properties via a unified hazard model has always been challenging and has yet to be attempted. Here, we develop a unified model to estimate landslide hazard at the slope unit level to address such gaps. We employed deep learning, combined with extreme‐value theory to analyze an inventory of 30 years of observed rainfall‐triggered landslides in Nepal and assess landslide hazard for multiple return periods. We also use our model to further explore landslide hazard for the same return periods under different climate change scenarios up to the end of the century. Our model performs excellently (with an accuracy of 0.78 and an area under the curve of 0.86) and can be used to model landslide hazard in a unified manner. Geomorphologically, we find that under climate change scenarios (SSP245 and SSP585), landslide hazard is likely to increase up to two times on average in the lower Himalayan regions (Siwalik and lower Himalayas; ≈110–3,500 m) while remaining the same in the middle Himalayan region (≈3,500–5,000 m) whilst decreasing slightly in the upper Himalayan region (≳5,000 m) areas. Plain Language Summary We implement a model capable of satisfying the standard definition of landslide hazard. Estimating landslide hazard requires assessing where landslides may occur, how frequently and how large they may be. Since the inception of such a concept, each component of the hazard has been modeled individually, with most of the scientific efforts being dedicated to the spatial and temporal components, thus neglecting the third intensity requirement. Here, we make use of a deep learning architecture to estimate “where” by solving a standard classification problem jointly. As for the intensity, we modeled the landslide area density for each slope in the selected Nepalese landscape. This is done by solving a problem typical of extreme statistics, still in a deep learning framework. Because extreme statistics is suitable to make long‐term predictions, that is, at return periods of a given process, we intrinsically addressed the temporal component as well. Thanks to this property, we generate landslide hazard simulations for this century, incorporating climate projections in our analyses. Key Points A Landslide hazard model is developed by combining deep learning and the extended generalized Pareto distribution from extreme value theory Thirty years of past observations are used to model the current landslide hazard scenario under different return periods of precipitation Landslide hazard is predicted for multiple return periods of climate change scenarios up to the end of the century