Dataset

Post-earthquake topography and landslide inventory from the 2015 Lefkada Mw6.5 earthquake, Greece

Clark, M.K., Zekkos, D. & Manousakis, J. Enabling 3D landslide event statistics using satellite and UAV-enabled topographic differencing. Landslides 22, 1041–1058 (2025). https://doi.org/10.1007/s10346-024-02374-x Landslides represent one of the most significant hazards during and in the immediate aftermath of an earthquake. Time series of high-resolution imagery from satellites and UAVs present new opportunities for quantifying volume inventories in terms of scale and resolution. Maps of elevation change can more accurately identify landslide source areas on steep slopes compared to imagery alone, distinguish deeper landsliding from shallow ravel, and reveal complex patterns that are not well approximated by simple landslide slip surface geometries. Statistical relationships of landslide inventories are also sensitive to aspects of the methodology, namely topographic resolution, accurate image registration, and estimation of 2D (plane) area versus 3D surface area. As we achieve resolution that surpasses ground-based field studies and removes the selection bias of focusing only on select well defined, deep landslides, new patterns of ground failure emerge to which mapping and statistical data interpretations will need to adapt. The purpose of this project was to document landscape change due to earthquake forcing. This research was a principal part of a USGS National Earthquake Hazards Reduction Program (NEHRP) grant (G17AP00088) and supported by grants from NASA and the NSF (80NSSC19K0948 & 80NSSC20K1032 (NASA Disasters Program), OPP-1043681, OPP-1559691 and OPP-1542736 (NSF). The data provided here are from field surveys conducted after the earthquake and satellite data analysis. Remotely-sensed stereo optical imagery collected shortly after the 2015 Lefkada Mw6.5 earthquake event in western Greece were used to create two post-event DSM surfaces using (1) Worldview-3 satellite images with the SETSM algorithm, and (2) UAV-based optical imagery using Structure from Motion (SfM). DEM differencing and optical image mapping of ~ 700 landslides was then used to produce a novel 3D volume dataset of landslides caused by the earthquake using topographic differencing methods. These analyses provide comparison of two different state-of-the-art technological mapping tools: Unmanned Aerial Vehicles (UAVs) equipped with optical cameras and high-resolution optical stereo-satellite imagery. Post-earthquake DSM and difference models are published along with the landslide inventory. We expect the data archived here to be of interest to academic researchers including geomorphologists and geotechnical engineers working on earthquake driven landsliding. Specific problems that these data can inform (amongst others) include the formation and runout dynamics of earthquake-triggered landsliding, coastal escarpment retreat and longshore coastal sediment transport. The data may also interest government agencies and other entities working on hazard and risk associated with earthquake-triggered landsliding. The data from this project were collected with the intention of capturing perishable information in the aftermath of a moderate-sized earthquake. These data are being made freely and openly available for use by any interested parties, including academic researchers and others. Potential uses could include (but are not limited to) change detection, evaluation of material properties that aid in the mechanistic understanding of landslide processes, model parametrization, including quantification of landslide trigger and runout. The data collected during this project also provide a baseline that can be used to evaluate future changes associated with land surface hazards and coastal escarpment retreat at the study location.

https://www.designsafe-ci.org/data/browser/public/designsafe.storage.published/PRJ-5724

http://schema.org/True

https://opendatacommons.org/licenses/by/

landslide

earthquake

runout

erosion

DSMs

topographic differencing

imagery-based topographic models

Person

Marin

Clark

Marin Clark

Person

Dimitrios

Zekkos

Dimitrios Zekkos

Dataset

Landscape change and landslide inventories from the 2015 Mw6.5 Lefkada Earthquake, Greece

Clark, M.K., Zekkos, D. & Manousakis, J. Enabling 3D landslide event statistics using satellite and UAV-enabled topographic differencing. Landslides 22, 1041–1058 (2025). https://doi.org/10.1007/s10346-024-02374-x This mission documents landscape change caused by the 2015 Mw6.5 Lefkada earthquake in western Greece. To capture perishable data in the aftermath, we combined satellite and UAV remote sensing with field surveys to map earthquake-triggered landslides and quantify topographic change. High-resolution WorldView-3 stereo imagery was processed with the SETSM algorithm to generate digital surface models (DSMs), while UAV optical imagery was processed with Structure-from-Motion (SfM) to create local DEMs. DEM differencing between pre- and post-event surfaces revealed elevation change and enabled mapping of ~700 landslides, producing a novel three-dimensional landslide volume inventory. Results show that topographic differencing provides clearer identification of source areas than imagery alone, distinguishes deep failures from shallow ravel, and captures complex patterns not represented by simple slip-surface models. Data products include WorldView-3 DEMs, UAV-derived DEMs, DEM difference rasters, a polygon inventory of landslides, and supporting peak ground acceleration (PGA) grids. These openly available datasets can be reused to inform hazard and risk assessments, evaluate landslide initiation and runout dynamics, improve change-detection methods, and provide baselines for future studies of earthquake-driven geomorphic processes and coastal retreat.

Earthquake-triggered landslides

Lefkada earthquake 2015

Western Greece

Digital Elevation Model (DEM)

DEM differencing

UAV photogrammetry

Structure from Motion (SfM)

Stereo satellite imagery

WorldView-3

Landslide volume inventory

1

https://opendatacommons.org/licenses/by/

https://www.designsafe-ci.org/data/browser/public/designsafe.storage.published/PRJ-5724/#detail-f50f72631-54a1-4bc7-9382-4b0613af88b8

Person

Marin

Clark

Marin Clark

Person

Dimitrios

Zekkos

Dimitrios Zekkos

Person

John

Manousakis

John Manousakis

Organization

University of Michigan (umich.edu)

Organization

University of California Berkeley (berkeley.edu)

https://doi.org/10.17603/ds2-jk9k-m202

Dataset

Landscape change and landslide inventories from the 2015 Mw6.5 Lefkada Earthquake, Greece

Clark, M.K., Zekkos, D. & Manousakis, J. Enabling 3D landslide event statistics using satellite and UAV-enabled topographic differencing. Landslides 22, 1041–1058 (2025). https://doi.org/10.1007/s10346-024-02374-x This mission documents landscape change caused by the 2015 Mw6.5 Lefkada earthquake in western Greece. To capture perishable data in the aftermath, we combined satellite and UAV remote sensing with field surveys to map earthquake-triggered landslides and quantify topographic change. High-resolution WorldView-3 stereo imagery was processed with the SETSM algorithm to generate digital surface models (DSMs), while UAV optical imagery was processed with Structure-from-Motion (SfM) to create local DEMs. DEM differencing between pre- and post-event surfaces revealed elevation change and enabled mapping of ~700 landslides, producing a novel three-dimensional landslide volume inventory. Results show that topographic differencing provides clearer identification of source areas than imagery alone, distinguishes deep failures from shallow ravel, and captures complex patterns not represented by simple slip-surface models. Data products include WorldView-3 DEMs, UAV-derived DEMs, DEM difference rasters, a polygon inventory of landslides, and supporting peak ground acceleration (PGA) grids. These openly available datasets can be reused to inform hazard and risk assessments, evaluate landslide initiation and runout dynamics, improve change-detection methods, and provide baselines for future studies of earthquake-driven geomorphic processes and coastal retreat.

Earthquake-triggered landslides

Lefkada earthquake 2015

Western Greece

Digital Elevation Model (DEM)

DEM differencing

UAV photogrammetry

Structure from Motion (SfM)

Stereo satellite imagery

WorldView-3

Landslide volume inventory

1

https://opendatacommons.org/licenses/by/

https://www.designsafe-ci.org/data/browser/public/designsafe.storage.published/PRJ-5724/#detail-f50f72631-54a1-4bc7-9382-4b0613af88b8

Person

Marin

Clark

Marin Clark

Person

Dimitrios

Zekkos

Dimitrios Zekkos

Person

John

Manousakis

John Manousakis

Organization

University of Michigan (umich.edu)

Organization

University of California Berkeley (berkeley.edu)

https://doi.org/10.17603/ds2-jk9k-m202