COSMO-SkyMed~® potentiality to identify crop-specific behavior and monitor phenological parameters

摘要

This work aims at investigating the capability of COSMO-SkyMed~® (CSK~®) constellation of Synthetic Aperture Radar (SAR) system to monitor the Leaf Area Index (LAI) of different crops. The experiment was conducted in the Marchfeld Region, an agricultural Austrian area, and focused on five crop species: sugar beet, soybean, potato, pea and corn. A linear regression analysis was carried out to assess the sensitivity of CSK~® backscattering coefficients to crops changes base on LAI values. CSK~® backscattering coefficients were averaged at a field scale (<σ~o_(dB)>) and were compared to the DEIMOS-1 derived values of estimated LAI. LAI were as well averaged over the corresponding fields (). CSK~® data acquired at three polarizations (HH, VV and VH), four incidence angles (23°, 33°, 40° and 57°) and at different pixel spacings (2.5 m and 10 m) were tested to assess whether spatial resolution may influence results at a field scale and to find the best combination of polarizations and CSK~® acquisition beams which indicate the highest sensitivity to crop LAI values. The preliminary results show that sugar beet can be well monitored (r = 0.72 - 0.80) by CSK~® by using any of the polarization acquisition modes, at moderate to shallow incidence angles (33° - 57°). Slightly weaker correlations were found, at VH polarization only, between CSK~® < σ~o_(dB)> and for potato (r = 0.65), pea (r = 0.65) and soybean (r = -0.83). Shallower view incidence angles seem to be preferable to steep ones in most cases. CSK~® backscattering coefficients were no sensitive at all to LAI changes for already developed corn fields.