Interpretation of Backscattering Mechanisms from Non-gaussian Correlated Randomly Rough Surfaces

• Published in: Journal of electromagnetic waves and applications 2006-01, Vol.20 (1), p.105-118
• Main Authors: Chen, K. S., Fung, A. K., Shi, J. C., Lee, H. W.
• Format: Article
• Language: English
• Subjects: Applied sciences; Diffraction, scattering, reflection; Exact sciences and technology; Radiocommunications; Radiowave propagation; Telecommunications; Telecommunications and information theory
• ISSN: 0920-5071; 1569-3937
• DOI: 10.1163/156939306775777404

In this study we analyzed backscattering mechanism using a non-Gaussian correlation function, called exponential-like, for rough surface. The corresponding surface spectrum properties were investigated. It contains an excessive amount of high frequency spectral components. But unlike exponential one, it has rms slopes and allows to obtain the desired rms slopes. New insight into the physical implications of mechanism behind the backscattering behavior was obtained using the advanced integral equation model (AIEM). It is known that backscattering at large incident angles is proportional to the presence of small scale roughness or the high-frequency spectral components of the surface spectrum. Similarly, the peaking at near normal incidence is proportional to the large scale roughness or the low frequency spectral components of the surface spectrum. However, by increasing frequency while keeping the small roughness scale, scattering cannot approach the geometric optics at large angles of incidence where small roughness scales dominate scattering. There is condition where VV tends to HH returns without approaching the geometric optics limit. This is where we increase the surface rms slope. In fact, the presence of roughness scales, small or comparable to the incident wavelength, implies that the geometric optics condition is not satisfied. When there is a significant amount of backscattering in the large incident angle region, it indicates presence of high frequency spectral components in the surface spectrum. To demonstrate the applicability of the proposed correlation function, comparison of backscattering coefficients between model predictions and insitu measurements was made. Well match was obtained in terms of level and angular trend.