抄録
We conducted a cross-calibration experiment on airborne L-band SAR systems (JAXA/CRL Pi-SAR and JPL AIRSAR) in the Tottori Sand Dune. This experiment forms a part of the Pacific Rim-2000 campaign conducted from July to October in 2000. Both Pi-SAR and AIRSAR acquired their data over the test site within two hours on October 4, 2000. The aircraft track altitudes were about 8000 m. Four rectangular trihedral corner reflectors with an 80 cm leg-length were deployed in order to estimate absolute calibration coefficients. We also deployed four dihedral corner reflectors with different rotation angles for polarimetric calibration. This report describes the results of calibration experiment for Pi-SAR and AIRSAR systems conducted on the Tottori Sand Dune.
The Pi-SAR distortion matrices for both transmitting and receiving systems were calculated by using the data containing one trihedral reflector and two dihedral reflectors with different rotation angles. Gain balance between HH and VV was not good this time because Pi-SAR had some hardware degradation. The absolute calibration coefficient for the HH channel has almost the same as in the previous calibration experiment in 1998. However, we found that the VV channel had degraded more than 3 dB. The improvements of gain balance and crosstalk between H and V channels were confirmed after the calibration using the extracted distortion matrices.
We compared Pi-SAR data with AIRSAR data at point targets as well as distributed targets. Relatively large differences were observed in low backscattering targets with VV and HV channels. These differences are presumably due to increased noise equivalent backscattering coefficients caused by the hardware degradation of Pi-SAR V system data chain.
