Analysis of the applicability of multi-temporal full polarimetric airborne L-band SAR scattering to paddy rice field mapping

Chinatsu Yonezawa, Manabu Watanabe

    Research output: Contribution to journalArticlepeer-review

    Abstract

    A Synthetic Aperture Radar (SAR) is an all-weather imaging system that is often used for mapping paddy rice fields and estimating the area. Fully polarimetric SAR is used to detect the microwave scattering property. In this study, a simple threshold analysis of fully polarimetric L-band SAR data was conducted to distinguish paddy rice fields from soybean and other fields. We analysed a set of ten airborne SAR L-band 2 (Pi-SAR-L2) images obtained during the paddy rice growing season (in June, August, and September) from 2012 to 2014 using polarimetric decomposition. Vector data for agricultural land use areas were overlaid on the analysed images and the mean value for each agricultural parcel computed. By quantitatively comparing our data with a reference dataset generated from optical sensor images, effective polarimetric parameters and the ideal observation season were revealed. Double bounce scattering and surface scattering component ratios, derived using a four-component decomposition algorithm, were key to extracting paddy rice fields when the plant stems are vertical with respect to the ground. The alpha angle was also an effective factor for extracting rice fields from an agricultural area. The data obtained during August show maximum agreement with the reference dataset of estimated paddy rice field areas.

    Original languageEnglish
    Pages (from-to)2500-2516
    Number of pages17
    JournalInternational Journal of Remote Sensing
    Volume41
    Issue number7
    DOIs
    Publication statusPublished - 2020 Apr 2

    ASJC Scopus subject areas

    • Earth and Planetary Sciences(all)

    Fingerprint

    Dive into the research topics of 'Analysis of the applicability of multi-temporal full polarimetric airborne L-band SAR scattering to paddy rice field mapping'. Together they form a unique fingerprint.

    Cite this