Landslide detection dataset: Sentinel-2 and topographic data for Yecheon (2023) and Gokseong (2020) testbeds in South Korea

Description

This dataset accompanies the manuscript: > Jimin Jang, Jun-Hyuk Jang, Tae-Hyuk Kwon(2026) A transferable, deterministic landslide detection > Framework using Sentinel-2 and topographic data for rapid inventory > Construction in forested regions. GIScience & Remote Sensing. It contains ground truth (GT) polygons, algorithmically detected polygons, source data behind selected figures, and the QGIS Python scripts used to produce the detection results for two South Korean testbeds: Yecheon (July 2023) and Gokseong (August 2020). dataset/ ├── README.md ├── GT_dataset/ # ground truth landslide polygons │ ├── GT_dataset_Yecheon.gpkg # 93 polygons (Yecheon, 2023) │ └── GT_dataset_Gokseong.gpkg # 61 polygons (Gokseong, 2020) ├── Detection_result/ # algorithmically detected polygons │ ├── Detection_result_Yecheon.gpkg # 160 polygons (Yecheon, 2023) │ └── Detection_result_Gokseong.gpkg # 96 polygons (Gokseong, 2020) ├── Detection_script/ # QGIS Python scripts that produced the detection results from Sentinel-2 imagery and DEM derivatives │ ├── LDscript_Yecheon.py │ └── LDscript_Gokseong.py └── Figure_data/ # source data behind selected figures ├── Fig5_dNDVI.tif # ΔNDVI raster, Yecheon testbed (Fig. 5a) ├── Fig5_dBSI.tif # ΔBSI raster, Yecheon testbed (Fig. 5b) ├── Fig6_GT_stat.csv # topographic statistics of GT polygons (Fig. 6) └── Fig8_Parameter_calib_optimiz.xlsx # parameter calibration results (Fig. 8)

Institutions

• Korea Advanced Institute of Science and Technology

  Daejeon, Daejeon

Categories

Licence