Dataset for: Pre-collapse spaceborne deformation monitoring of the Kakhovka dam, Ukraine, from 2017 to 2023

Abstract

The Kakhovka Dam, situated on the Dnieper River in Ukraine, suffered a collapse on June 6, 2023, resulting in downstream inundation. In this dataset, we provide Sentinel-1 geocoded images acquired both before and after the incident—specifically, on June 1, 2023, and June 13, 2023, for the ascending geometry, and on May 28, 2023, and June 9, 2023, for the descending geometry. Additionally, we include the 50 cm resolution UMBRA descending geocoded image acquired on June 6, 2023, to identify parts of the dam that were breached. We then used a stack of ascending and descending Sentinel-1 A/B interferometric images spanning from 2015 to 2023 to generate a pre-collapse line of sight time-series of surface deformation of the dam. Our analysis reveals distinct deformation signals indicating pre-existing potential ongoing damage mechanisms on the southern part of the dam.

Table 1: Acquisition attributes of both the ascending and descending Sentinel1 images, used for the MT-InSAR process.

Description of the data and file structure

This dataset contains the Sentinel-1 and UMBRA geocoded images as well as the Sentinel-1 MT-InSAR results over Kakhovka dam in Ukraine.

GRDs folder contains the Sentinel-1 pre and after-the-dam collapse, geocoded images as well as an UMBRA image acquired after the collapse of the dam.

LOS_DCS.csv contains the MT-InSAR results of processing Sentinel-1 descending images, acquired from 2015 to 2023, deformation is expressed in millimeters. 

LOS_ACS.csv contains the MT-InSAR results of processing Sentinel1 ascending images, acquired from 2017 to 2023, deformation is expressed in millimeters. 

WaterLevel.xlsx contains the water-level data from Hydroweb website in meter.

AllTemp.xlsx contains the weather temperature from Visual Crossing website in celsius.

Funding

The Commercial Smallsat Data Scientific Analysis Program of NASA: NNH22ZDA001N-CSDSA

The Decadal Survey Incubation Program: Science and Technology: NNH21ZDA001N-DSI

The Vidi project InStruct, project number 18912, financed by the Dutch Research Council (NWO)