[H-GM02] Geomorphology

[HGM02-P02] Analysis of channel-head migration using pre-event airborne LiDAR data and post-event UAV-LiDAR data: A case of a heavy rainfall event in July, 2018

*Kenta Saito¹, Tsuyoshi Hattanji², Takuro Ogura³, Takahisa Furuichi⁴, Yasushi Tanaka⁵, Shoji Doshida⁶ (1.Graduate School of Science and Technology, University of Tsukuba, 2.Faculty of Life and Environmental Sciences, University of Tsukuba, 3.Graduate School of Education, Hyogo University of Teacher Education, 4.Forestry and Forest Products Research Institute, 5.Faculty of Letters, Komazawa University, 6.National Research Institute of Fire and Disaster)

Keywords:sediment disaster, shallow landslide, source area

A channel head is defined as the upstream-most point of a channel in a mountain valley. It has been inferred that a channel head migrates upstream immediately after heavy rainfall event due to shallow landslide, and that a channel head can also migrate downstream after several hundred years after the last landslide event. Understanding the migration of channel head is a key to solve prediction of shallow landslide as well as landscape evolution in mountain. High-resolution DEM is required to identify micro-topography including headwater channels and channel heads in a mountain. The recent spread of UAV-LiDAR has enabled us to acquire high-density point cloud data in inaccessible areas, and to compare topographic changes from multiple periods. The purpose of this study is to analyze the changes in the location of channel heads before and after the heavy rainfall event in July, 2018, using pre-event airborne LiDAR data and post-event UAV-LiDAR data. The survey area is located around Mt. Ege-san in Hiroshima Prefecture, Japan, where many shallow landslides occurred due to the heavy rainfall event in July, 2018. Location of pre-event channel heads were identified from slope map and contour map from the airborne LiDAR DEM obtained in 2009. Post-event DEMs with resolutions of 1 m and 0.5 m are created from UAV-LiDAR data obtained in November 2022. Location of post-event channel heads are identified from slope maps and contour maps from these data. There were 33 channel heads in a survey area of UAV-LiDAR. Shallow landslide occurred in source area of 7 of the 33 channel heads, and these channel heads migrated upstream from the pre-event position. Pre-event channel heads were classified into two types based on the occurrence or non-occurrence of shallow landslides. Topographic attributes of the channel heads were different, i.e., channel heads without shallow landslides have smaller source-area size than channel heads with shallow landslides. This result implies that channel heads with smaller source areas has lower probability of shallow landslides.