Advancing landslide hazard assessment around the world

Landslide disasters continue to be pervasive globally and advances in landslide hazard assessment are needed to inform response and risk reduction. To address this challenge, NASA has developed the Landslide Hazard Assessment for Situational Awareness (LHASA) system, which has been implemented at regional and global scales. LHASA integrates satellite precipitation information within a machine learning model leveraging a suite of satellite and ground-based products to indicate where and when landslide hazard may be elevated in near real-time (NRT). Several improvements have been made to LHASA, including the use of global forecast data for landslide prediction. These forecasts are most accurate for the first two days and for major landslide events such as those caused by tropical storms. In addition, a preliminary global post-fire debris flow model has been developed to identify watersheds with the highest probability of debris flow occurrence following wildfire in NRT.LHASA runs routinely at the NASA Goddard Space Flight Center and the Pacific Disaster Center.

Reliable landslide inventories are key to evaluating LHASA. The Semi-Automatic Landslide Detection (SALaD) system has been employed in both local and regional settings in support of rapid response efforts. SALaD inventories were generated within Nepal's Arun River basin for several years and formed the basis for landslide susceptibility mapping and hazard assessment in advance of anticipated regional hydropower projects. To address the potential for landslide disasters in Southeast Asia, SALaD was used to generate inventories for the region while the LHASA system was adapted for use by stakeholders in the Lower Mekong Region. Finally, the historical performance of LHASA over Puerto Rico was evaluated to inform a multi-agency workshop focused on advancing hazard assessment and disaster response over the island. The suite of tools encompassing LHASA and SALaD have been foundational in support of several major landslide events and enabled rapid mapping of landslides, landslide densities, and/or assessment of related products such as numerical weather forecasts. This work advances landslide hazard assessment through open data and modeling frameworks. This presentation will highlight both new capabilities and future paths of this research.