Modeling L-band radar backscatter of Alaskan boreal forest

• Published in: IEEE transactions on geoscience and remote sensing 1993-11, Vol.31 (6), p.1146-1154
• Main Authors: Wang, Y., Day, J.L., Davis, F.W., Melack, J.M.
• Format: Article
• Language: English
• Subjects: Backscatter; Biomedical monitoring; Dielectrics; Earth sciences; Earth, ocean, space; Exact sciences and technology; External geophysics; Geophysics. Techniques, methods, instrumentation and models; L-band; Polarization; Predictive models; Remote monitoring; Soils; Spaceborne radar; Surficial geology; Synthetic aperture radar; Testing
• ISSN: 0196-2892; 1558-0644
• DOI: 10.1109/36.317448

Synthetic aperture radar (SAR) data were acquired over Bonanza Creek Experimental Forest (Alaska) in March 1988 under thawed and frozen conditions. For five stands analyzed, L-band backscatter at 42 degrees -45 degrees incidence angle was 2.7-6.9 dB smaller under frozen than under thawed conditions for white spruce and balsam poplar, with the largest difference at HV and the smallest at HH polarization. The differences were smaller for a stand of small black spruce. The VV-HH phase differences observed by SAR were approximately=0 degrees for all the stands. Ground data were used to parameterize the Santa Barbara canopy backscatter model. For the white spruce and balsam poplar stands under thawed conditions, simulations agreed with the SAR data within the calibration uncertainty. The model underestimated the HH, HV, and VV backscatter for all five stands under frozen conditions, and for the black spruce stand under thawed conditions. The modeled VV-HH phase differences were close to 0 degrees for all the stands except the black spruce stand. The discrepancies in model predictions of backscatter and phase difference were attributed to inadequate surface backscatter modeling. Model results supported the hypothesis that the weaker backscatter from frozen stands was because of the smaller dielectric constant of the frozen trees.< >