TY - JOUR

T1 - Extracting sinkhole features from time-series of TerraSAR-X/TanDEM-X data

AU - Vajedian, Sanaz

AU - Motagh, Mahdi

N1 - Funding information: TDX data are copyright German Aerospace Agency (DLR) and were provided under proposal number motagh_XTI_LAND6959. This work was partially supported by the Initiative and Networking Fund of the Helmholtz Association in the frame of the Helmholtz Alliance’s ‘‘Remote Sensing and Earth System Dynamics”. We are grateful to Paolo Riccardi, Alessio Cantone and Paolo Pasquali from Sarmap for their technical help and support of Sanaz Vajedian's visit to their company. We thank Ahmad Hojati for his comments and Bahman Akbari for his assistance during field survey. We also thank two anonymous reviewers for a number of suggestions that helped us improve the paper.

PY - 2019/4

Y1 - 2019/4

N2 - Sinkholes are significant geologic hazards that are mainly formed in water-soluble carbonate bedrocks such as limestone, dolomite or gypsum. Sinkhole formation causes the surface to subside or collapse suddenly without any prior warning, and therefore can lead to extensive damage and even loss of life and property. Delineating sinkholes is important for understanding hydrological processes and mitigating geological hazards in karst areas. The recent development in deriving high-resolution digital elevation models from space missions such as TerraSAR-X/TanDEM-X (TSX/TDX) enables us to delineate and analyze geomorphologic features and landscape structures at small scale (up to 2 m). In this study we use time-series of TSX/TDX data and develop an adaptive sinkhole-analysis method using interferometry observations. A wavelet-based refinement approach is implemented on interferomeric processing to reduce the baseline bias effects and align the interferometrically-derived DEMs. The multi-temporal DEMs are then successfully stacked using Canonical Correlation Analysis (CCA) to reconstruct a higher quality DEM. Finally, feature extraction using watershed algorithm is applied to precisely delineate geomorphometric characteristics of the sinkholes. Five TSX/TDX images are selected to evaluate the performance of our approach for sinkholes in Hamedan, West Iran. Results show that applying our methodology on high-resolution TSX/TDX data from different geometries and time periods enables us to effectively distinguish sinkholes from other depression features of the basin. Different TSX/TDX pairs produce consistent results for diameter and depth of sinkholes with the standard deviation of approximately 1 m, in agreement with field observations.

AB - Sinkholes are significant geologic hazards that are mainly formed in water-soluble carbonate bedrocks such as limestone, dolomite or gypsum. Sinkhole formation causes the surface to subside or collapse suddenly without any prior warning, and therefore can lead to extensive damage and even loss of life and property. Delineating sinkholes is important for understanding hydrological processes and mitigating geological hazards in karst areas. The recent development in deriving high-resolution digital elevation models from space missions such as TerraSAR-X/TanDEM-X (TSX/TDX) enables us to delineate and analyze geomorphologic features and landscape structures at small scale (up to 2 m). In this study we use time-series of TSX/TDX data and develop an adaptive sinkhole-analysis method using interferometry observations. A wavelet-based refinement approach is implemented on interferomeric processing to reduce the baseline bias effects and align the interferometrically-derived DEMs. The multi-temporal DEMs are then successfully stacked using Canonical Correlation Analysis (CCA) to reconstruct a higher quality DEM. Finally, feature extraction using watershed algorithm is applied to precisely delineate geomorphometric characteristics of the sinkholes. Five TSX/TDX images are selected to evaluate the performance of our approach for sinkholes in Hamedan, West Iran. Results show that applying our methodology on high-resolution TSX/TDX data from different geometries and time periods enables us to effectively distinguish sinkholes from other depression features of the basin. Different TSX/TDX pairs produce consistent results for diameter and depth of sinkholes with the standard deviation of approximately 1 m, in agreement with field observations.

KW - Bistatic interferometry

KW - CCA analysis

KW - Sinkhole

KW - TanDEM-X

KW - Wavelet decomposition

UR - http://www.scopus.com/inward/record.url?scp=85062476296&partnerID=8YFLogxK

U2 - 10.1016/j.isprsjprs.2019.02.016

DO - 10.1016/j.isprsjprs.2019.02.016

M3 - Article

AN - SCOPUS:85062476296

VL - 150

SP - 274

EP - 284

JO - ISPRS Journal of Photogrammetry and Remote Sensing

JF - ISPRS Journal of Photogrammetry and Remote Sensing

SN - 0924-2716

ER -