Significance of radar remote sensed imagery for agricultural applications

摘要

Remotely sensed optical image data acquired by airborne or satellite sensors provide useful information about crop discrimination as well as infield crop variability. It has been demonstrated that optical remote sensing data is applicable to estimatecrop biomass variability for developing management strategies. Due to the weather dependency of this technique, data availability is not guaranteed when it is needed. A solution to this problem might be radar technology which can be acquired at any time,which means at all relevant growth stages. Synthetic aperture radar (SAR) backscattering provides information about canopy structure and water content. The insufficient understanding of SAR backscatter mechanisms according to agricultural soil and plantconditions is the biggest gap of SAR data usage in agriculture at present time. The objective of the presented study was to improve the understanding of radar backscattering of agricultural crops and to show the potential and restrictions of radar remote sensing in agriculture. Research was performed during the vegetation periods of 2001 and 2002 in an agricultural dominated area of Mecklenburg-Western Pomera-nia in northern Germany (13.4° E; 53.9° N). Multi-dimensional airborne X- and L-band SAR-data were acquired by the German E-SAR (DLR) with l m ground resolution. The results indicate that different crops can be discriminated and infield variability can be located clearly by radar imagery. Further single radar bands and several different band combinations were tested for determining infield biomass variability in winter wheat. Therefore a correlation analytical and a discriminant analytical approach were chosen. It turned out, that backscatter of horizontal polarized X-band (X-HH) is most suitable, but relationships were improved by using band combinations. Likewise the generalized discriminant analysis revealed better relationships than the correlation analysis. Despite the difficulties of understanding radar backscattering, radar imagery provides valuable information about infield spatial crop variability, even if the acquisition of optical data is limited by unfavorable weather conditions.