Monitoring of Agricultural Grasslands With Time Series of X-Band Repeat-Pass Interferometric SAR

Abstract

This study evaluates the potential of X-band interferometry for monitoring of agricultural grasslands. Time series of HH-polarization COSMO-SkyMed 1-day repeat-pass interferometric SAR (InSAR) pairs is analyzed in regard to detecting mowing events, and assessing vegetation height and biomass on grasslands. The time series of four InSAR pairs was analyzed in regard to the ground reference data collected during an extensive campaign covering 11 agricultural grasslands. The calculated temporal interferometric coherence was found to be inversely correlated to the vegetation height and wet above-ground biomass. It was found that grass removal increases the coherence magnitude indicating a potential use of this parameter for the detection of mowing. However, precipitation and farming activity between the acquisitions interfere with this effect. Temporal coherence was expressed as a function of the vegetation height through the random motion of scatterers in the vegetation layer. For vegetation height limited to the range between 0 and 1 m, a very strong correlation between the grass height and the linearised temporal coherence was found, with a coefficient r = 0.81. No significant correlation was found between the backscattering coefficient and the wet above-ground biomass as well as the height of grass. However, a strong negative correlation was found between the backscattering coefficient and the measured soil moisture.