An investigation into the capability of hyperspectral data to detect buried pipeline remediation induced vegetation stress (2005)

Author(s): White DC, Williams M, Barr SL

    Abstract: The preliminary results of a project investigating the suitability of hyperspectral imagery for monitoring pipeline remediation are presented. CASI-2 imagery and field spectroscopy have been acquired along a 9 km stretch of buried pipeline in Aberdeenshire encompassing a range of arable crop types. The feasibility of identifying areas of vulnerability for monitoring buried pipeline remediation from VIS-NIR and SWIR field spectrometry are investigated. Vegetation stress associated with pipeline remediation was investigated through analysis of the position, shape and magnitude of spectral reflectance and absorption features from the VIS-NIR to the SWIR. Results suggest that changes in reflectance of 3 - 20 % between spectral stress features occur at the sub-metre level for spring oats and barley at early phenological stages. First derivative spectra of stress relative to control sites exhibit a less pronounced second peak in red edge at 732 nm. Peaks at 698 and 718 nm are relatively more pronounced with first derivative values of 0.07 to 0.08 and 0.07, higher than the double peak in the red edge, respectively. The trough at 572 nm is up to three times shallower for stress relative to control spectra. SWIR features are less consistent but notable features are apparent in troughs around 1130 to 1150 nm being up to twice as shallow.

      • Date: 6-9 September 2005
      • Conference Name: Annual Conference of the Remote Sensing and Photogrammetry Society
      • Publication type: Conference Proceedings (inc. abstract)
      • Bibliographic status: Unknown

        Keywords: airborne remote sensing; CASI; pipeline remediation


        Dr Stuart Barr
        Senior Lecturer in Geographic Information Science