ESTIMATION OF POLARIZATION ERRORS FROM COVARIANCE MATRICES OF CSU-CHILL RADAR DATA

摘要

The covariance matrix of weather targets can be measured by the CSU-CHILL radar which alternately transmits pulses of H (horizontal) and V (vertical) polarized energy and receives simultaneously both H and V polarized signals. The covariance matrix is a function of not only the scattering process but is also contaminated by polarization errors of the radar. If the covariance matrices are well calibrated it is possible to estimate these polarization errors from data. The expected value of the polarization errors is quite small and thus the copolar measurends will be negligibly affected. However, the crosspolar measurands can be greatly affected even by polarization errors of a few tenths of a degree (Hubbert et al., 1999). Thus knowledge of polarization errors is important to the interpretation of radar variables such as LDR (liner depolarization ratio) and ρ_(hh,vh), the co-to-cross correlation coefficient.