Landslides are a destructive natural disaster causing damage and loss of life each year. Researchers from Rochester Institute of Technology introduced a new paradigm for studying landslide shapes and failure types, aiding in predicting landslides and risk evaluations. Ph.D. student Kamal Rana and assistant professor Nishant Malik were lead authors on a paper in Nature Communications, along with Kushanav Bhuyan. By using aerial view and elevation data combined with machine learning, researchers achieved 80-94 percent accuracy in identifying landslide movements globally, introducing a method of examining slides, flows, and fails.
Studying landslides worldwide, including the 2008 disaster in Beichuan, China, researchers developed a new paradigm to understand their movements and failure types. The algorithm does not predict landslides but provides information on the causes and mechanisms behind them. Studied locations included Italy, the United States, Denmark, Turkey, and China, confirming the strength of the findings in diverse regions and climates. The real-world application of this research has a personal impact for lead author Kamal Rana, who hails from the Himalayan region of India and has witnessed the devastating effects of landslides blocking roads and disrupting daily life.
The hope is that a deeper understanding of landslide failure movements will aid in predicting deadly events and enhancing the accuracy of hazard and risk assessment models. This research can potentially save lives and reduce damage by providing valuable information for those working on predicting landslides. Co-authors of the paper include researchers from the University of Potsdam and the University of Padova, among others, contributing to a comprehensive study on landslide shapes and failure types globally.
By incorporating aerial view and elevation data with machine learning, researchers have achieved high accuracy in identifying landslide movements, which will aid in predicting landslides and risk assessments globally. The study introduces a novel method of examining different types of landslides, such as slides, flows, and fails, to better understand the causes and mechanisms behind them. The success of the algorithm in diverse locations around the world highlights the potential impact of this research on improving landslide prediction and risk evaluation models.
Studying landslides in various countries, including China, Italy, Turkey, Denmark, and the United States, researchers have developed a new paradigm to understand landslide movements and failure types. The algorithm does not predict landslides but provides valuable information for those working on predicting landslides to understand the causes and mechanisms behind them. The strength of the findings across diverse regions and climates confirms the potential impact of this research in improving hazard and risk assessment models globally.
The personal impact of this research on lead author Kamal Rana, who has witnessed the devastating effects of landslides in the Himalayan region of India, highlights the importance of predicting landslides accurately to save lives and reduce damage. By providing a deeper understanding of landslide failure movements, this research aims to enhance the accuracy and reliability of hazard and risk assessment models, ultimately helping to predict deadly events more effectively. Co-authored by researchers from various universities, this study is a comprehensive effort to improve the prediction and evaluation of landslides globally.
