EVALUATION OF CALIBRATION METHODS FOR A HELICOPTER-BORNE MICROWAVE SCATTEROMETER

Citation
M. Benallegue et al., EVALUATION OF CALIBRATION METHODS FOR A HELICOPTER-BORNE MICROWAVE SCATTEROMETER, International journal of remote sensing, 16(2), 1995, pp. 217-226
Citations number
12
Categorie Soggetti
Photographic Tecnology","Remote Sensing
ISSN journal
01431161
Volume
16
Issue
2
Year of publication
1995
Pages
217 - 226
Database
ISI
SICI code
0143-1161(1995)16:2<217:EOCMFA>2.0.ZU;2-1
Abstract
The French frequency modulated continuous waves (FMCW) scatterometer E RASME mounted on small helicopter or aircraft has been designed as dua lfrequency (C and X bands) and dualpolarization (HH, VV) to investigat e simultaneously the vegetation and the soil responses in radar backsc attering. It is operated as a forward looking radar with a large eleva tion beamwidth (+/- 10 degrees at 3 db) to observe easily the same sur face target over a large range of incidence angles during a single fli ght. By this ability, ERASME is a complementary research tool for inte rcalibration of airborne and spaceborne imaging Synthetic Aperture Rad ars like Radarsat and ASAR and has to be well calibrated in every conf iguration, both absolutely and relatively for comparisons at different incidence angles. This paper evaluates different calibration methods to be applied to such an instrument. Absolute calibration within 1 dB is easily obtained by external calibration using metallic corner refle ctors. But this method remains insufficient to get the antenna elevati on aperture which is essential on natural distributed targets for ante nna pattern correction, due to the severe constraint of a narrow azimu thal beam and flight parameters (pitch, roll, altitude) varying quickl y in time and range. The external calibration is strongly improved by using a statistical analysis of data obtained over natural targets whi ch analyses the correlation between the processed data and the recorde d flight parameters. This method appears promising, but its applicatio n on natural targets with random variations need specific statistical properties of the data set. It is operative for high antenna setting ( here 38 degrees incidence angle) and mostly over bare soils, with low sigma(0) variances and sigma(2) correlation length of the order of the correlation length of pitch. It provides the aperture range around th e antenna axis and an accuracy of 0.5dB upon sigma(0) is achieved prov iding the antenna pattern correction are done.