Multitemporal InSAR in land-cover and vegetation mapping

Marcus Engdahl

    Research output: ThesisDoctoral ThesisCollection of Articles

    Abstract

    Synthetic Aperture Radar (SAR) is an active microwave instrument that has a number of favourable characteristics, one of which is its independence from lighting conditions or cloud cover, which make it an indispensable instrument for earth observation from space. Classification of land cover and vegetation mapping are some of the major uses of SAR in environmental monitoring, and in order to reach the full potential of the SAR instrument, a coherent technique called SAR interferometry (InSAR) should be utilised. The so-called repeat-pass InSAR-techniques are possible only if the satellite is on a repeating orbit with a relatively short repeat-period (up to two weeks at C-band). This research focuses on multitemporal repeat-pass InSAR datasets from the Tandem- mission of the ERS-1 and ERS-2 satellites, which provided a time-series of InSAR-acquisitions with a 24-hour temporal baseline, a baseline which is very favourable to land-cover and vegetation mapping at C-band. The main research topic for this thesis is the assessment of the information content of a multitemporal InSAR time-series from an applications perspective. The work has concentrated on three application domains - land cover classification, stem volume estimation in boreal forests and the estimation of agricultural crop heights. This study demonstrates that multitemporal ERS-1/2 Tandem InSAR data is clearly superior to non-interferometric (intensity-only) data, and that the applications-potential of the data is very high for land cover classification, high to very high for boreal forest stem volume estimation and moderate to high for the estimation of agricultural crop heights. During the study novel methods were developed for both land cover classification and boreal forest stem volume estimation, as well as for their combination. An outline for an operational system for land cover classification and vegetation using multitemporal InSAR data is presented. The methods developed in this thesis can be utilised with data from the C-band ESA Sentinel-1 constellation, whose 12- and 6-day repeat InSAR data can be expected to make a large impact in both land cover classification and vegetation mapping.
    Translated title of the contributionMonen ajankohdan SAR-tutkainterferometria maankäytön luokittelussa ja kasvillisuuden kaukokartoituksessa
    Original languageEnglish
    QualificationDoctor's degree
    Awarding Institution
    • Aalto University
    Supervisors/Advisors
    • Hallikainen, Martti, Supervising Professor
    Publisher
    Print ISBNs978-952-60-5415-5
    Electronic ISBNs978-952-60-5416-2
    Publication statusPublished - 2013
    MoE publication typeG5 Doctoral dissertation (article)

    Keywords

    • remote sensing
    • earth observation
    • radar
    • synthetic aperture radar
    • SAR
    • SAR interferometry
    • InSAR
    • multitemporal analysis
    • multitemporal InSAR
    • land-cover classification
    • crop height
    • boreal forest stem volume estimation
    • vegetation mapping

    Fingerprint

    Dive into the research topics of 'Multitemporal InSAR in land-cover and vegetation mapping'. Together they form a unique fingerprint.

    Cite this