The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Download
Publications Copernicus
Download
Citation
Articles | Volume XLIII-B3-2020
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLIII-B3-2020, 739–744, 2020
https://doi.org/10.5194/isprs-archives-XLIII-B3-2020-739-2020
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLIII-B3-2020, 739–744, 2020
https://doi.org/10.5194/isprs-archives-XLIII-B3-2020-739-2020

  21 Aug 2020

21 Aug 2020

DETECTION OF LAND COVER DISPLACEMENTS THROUGH TIME-SERIES ANALYSIS OF MULTISPECTRAL SATELLITE IMAGERY: APPLICATION TO DESERT

D. Oxoli1, M. A. Brovelli1, D. Frizzi2, and S. Martinati2 D. Oxoli et al.
  • 1Department of Civil and Environmental Engineering, Politecnico di Milano, Milan, Italy
  • 2Saipem S.p.A., Milan, Italy

Keywords: Multispectral Imagery, Multitemporal Analysis, Land Cover Displacements, Sand Dunes, Open Data, Free and Open Source Software

Abstract. Detection of changes occurring on Earth surface has become an established practice in remote sensing science since the availability of high-resolution, global coverage, and multitemporal satellite imagery that has constantly increased during the last decades. Meanwhile, the open data policies embraced by some of the principal Earth Observation programs have boosted the spread of such analyses. This asset has attracted also the interest of the private sector which is developing strategies to exploit the potential of these data. In view of the above, we present an experimental procedure to investigate land cover displacements through an application on open multispectral imagery from the Landsat 8 and Sentinel-2 missions for desert sand dunes movements analysis. While most of the change detection techniques focus on locating changes and describing them by means of variation in the pixel spectral responses, the proposed technique aims at describing spatial and temporal patterns of displacements (i.e. directions and magnitude) applying cross-correlation analysis on a multitemporal images stack. Results of the proposed analysis are critical to a number of construction engineering operations that require sand mitigation planning and continuous site monitoring to prevent windblown sand interactions with infrastructures in the desert environment. An overview of the preliminary results is presented together with an extensive discussion on further improvements requested by the procedure. Computational steps leverage exclusively open data and free and open source GIS software thus providing large rooms to empower, replicate and improve thereof.