March 26, 2024
Identifying recurrent and persistent landslides using satellite imagery and deep learning: A 30-year analysis of the Himalaya
Tzu-Hsin Karen Chen, Mark E. Kincey, Nick J. Rosser, Karen C. Seto, Identifying recurrent and persistent landslides using satellite imagery and deep learning: A 30-year analysis of the Himalaya, Science of The Total Environment, Volume 922, 2024, 171161, ISSN 0048-9697, https://doi.org/10.1016/j.scitotenv.2024.171161.
Abstract
This paper presents a remote sensing-based method to efficiently generate multi-temporal landslide inventories and identify recurrent and persistent landslides. We used free data from Landsat, nighttime lights, digital elevation models, and a convolutional neural network model to develop the first multi-decadal inventory of landslides across the Himalaya, spanning from 1992 to 2021. The model successfully delineated >265,000 landslides, accurately identifying 83 % of manually mapped landslide areas and 94 % of reported landslide events in the region. Surprisingly, only 14 % of landslide areas each year were first occurrences, 55–83 % of landslide areas were persistent and 3–24 % had reactivated. On average, a landslide-affected pixel persisted for 4.7 years before recovery, a duration shorter than findings from small-scale studies following a major earthquake event. Among the recovered areas, 50 % of them experienced recurrent landslides after an average of five years. In fact, 22 % of landslide areas in the Himalaya experienced at least three episodes of landslides within 30 years. Disparities in landslide persistence across the Himalaya were pronounced, with an average recovery time of 6 years for Western India and Nepal, compared to 3 years for Bhutan and Eastern India. Slope and elevation emerged as significant controls of persistent and recurrent landslides. Road construction, afforestation policies, and seismic and monsoon activities were related to changes in landslide patterns in the Himalaya.
Keywords
Landslide inventory; Landslide evolution; Vegetation recovery; Multi-temporalSpatiotemporal analysis; Machine learning
