Reconstructing and assessing the lahar hazard at Changbaishan: insights from open-access numerical models

Abstract

Lahars at Changbaishan region are understudied despite being a major hazard due to its extensive river systems and large crater lake. Previous studies modelled the lahar hazard at Changbaishan under various scenarios, mostly using LaharZ due to the lack of field data. With new field data available, this study benchmarks LaharZ, Titan2D and LaharFlow by reconstructing the Millennium Eruption (ME) lahars before applying them for lahar hazard assessment at Changbaishan using model inputs informed by LaharDB, a new global lahar database. LaharDB builds upon an earlier unpublished database, and was reviewed to address various issues to improve its reliability and updated with recent data, before using it to constrain and infer model inputs. To reconstruct the ME lahars, model inputs were incrementally tested to determine the best-fit between model outputs and field extent. LaharZ best reconstructs the ME lahars; LaharFlow has the poorest reconstruction. The models provide insights into various aspects of lahars that are useful for land-use planning. For hazard assessment, scenarios defined by input volume, triggering mechanism and source location were modelled. The resulting runouts for crater lake, ice melt and rainfall scenarios are 2 – 65, 12 – 74 and 1 – 17 km respectively; the inundation areas are 1 – 683, 3 – 1137 and 0.5 – 17 km^2 respectively. The overall cropland, population, road, tree cover and built-up exposures are 0 – 22 km^2, 0 – 26,000 people, 0 – 588 km, 0.3 – 715 km^2 and 0 – 8 km^2 respectively. Careful land-use and hazard planning are important for minimising future lahar hazard and exposure.