Time-of-failure prediction of the Achoma landslide, Peru, from high frequency Planetscope satellites

Introduction

This repository contains the data used for the study of the slope instability of Achoma, Peru, described in Lacroix et al. (submitted). Specifically, the repository contains a time series of horizontal ground displacements, obtained from high frequency PlanetScope satellite between 2017 and 2020. It also contains two Digital Elevation Models, one from before the Achoma failure obtained with Pléaides stero images, and the other from just after the Achoma failure obtained with drone imagery.

The data and methods used for the elaboration of this data repository are described in detail in Lacroix et al. (submitted). In this repository we also provide a short summary and overview of the data and methods used.

Data

A total of 79 PlanetScope scenes were used to produce the time series of horizontal horizontal ground displacements maps. Table 1 provides an overview of these data.

Table1: Data used for the creation of this repository

Application

Platforms

Acquisition dates

Pre-failure DEM

Pléiades

2017/05/13

Post-failure DEM

Drone

2020/06/19

Horizontal ground displacement

PlanetScope

79 scenes from 2017/11/27 to 2020/06/17

Methods

The horizontal ground displacement maps, both along the NS and the EW directions (file names NSxxxxxxxx.tif and Ewxxxxxxxx.tif, where xxxxxxxx is the date in the format yyyymmdd) were created using the offset tracking methodology described in Bontemps et al. (2018), consisting of: (1) correlation of all the pairs of images using Mic-Mac (Rupnik et al., 2017), (2) masking the low correlation coefficient values (CC<0.7), (3) mosaicking correction, similar to stripe corrections (Bontemps et al., 2018), that we obtained by subtracting the median value of the stacked profile in the along-stripe direction, taking into account only stable areas, (4) least square inversion of the redundant system per pixel, weighted by the time separation between pairs (Bontemps et al., 2018), (5) correction of illumination effects (Lacroix et al., 2019), based on the 2 years of data between November 2017 and December 2019.

The pre-failure DEM was computed from Ames Stereo Pipeline (Shean et al. 2016) and the methodology developed in (Lacroix, 2016) applied to the Pléiades stereo images (file name DEM_20170513_shifted_vertical2.tif ).

The post-failure DEM was processed using the Structure from Motion-Multi View Stereo (SfM-MVS) methodology with the Agisoft Metashape Professional 1.5.5 software applied on 1824 pictures taken from the drone (file name Achoma_DEM_2020.06.20_UTM19S_50cm.tif ).

Acknowledgements

P.L. acknowledge the support from the French Space Agency (CNES) through the TOSCA, PNTS, and ISIS programs.

Dataset attribution

This dataset is licensed under a Creative Commons CC BY 4.0 International License.

Dataset Citation

Lacroix, P., Huanca, J., Angel, L., Taipe, E.: Data Repository: Time-of-failure prediction of the Achoma landslide, Peru, from high frequency Planetscope satellites. Dataset distributed on Zenodo: 10.5281/zenodo.7866962