[Influence of surface roughness on degree of polarization of biotite plagioclase gneiss varying with viewing angle].

Polarized reflectance is influenced by such factors as its physical and chemical properties, the viewing geometry composed of light incident zenith, viewing zenith and viewing azimuth relative to light incidence, surface roughness and texture, surface density, detection wavelengths, polarization phase angle and so on. In the present paper, the influence of surface roughness on the degree of polarization (DOP) of biotite plagioclase gneiss varying with viewing angle was inquired and analyzed quantitatively. The polarized spectra were measured by ASD FS3 spectrometer on the goniometer located in Northeast Normal University. When the incident zenith angle was fixed at 50 degrees, it was showed that on the rock surfaces with different roughness, in the specular reflection direction, the DOP spectrum within 350-2500 nm increased to the highest value first, and then began to decline varying with viewing zenith angle from 0 degree to 80 degrees. The characterized band (520 +/- 10) nm was picked out for further analysis. The correlation analysis between the peak DOP value of zenith and surface roughness showed that they are in a power function relationship, with the regression equation: y = 0.604x(-0.297), R2 = 0.985 4. The correlation model of the angle where the peak is in and the surface roughness is y = 3.4194x + 51.584, y < 90 degrees , R2 = 0.8177. With the detecting azimuth farther away from 180 degrees azimuth where the maximum DOP exists, the DOP lowers gradually and tends to 0. In the detection azimuth 180 dgrees , the correlation analysis between the peak values of DOP on the (520 =/- 10) nm band for five rocks and their surface roughness indicates a power function, with the regression equation being y = 0.5822x(-0.333), R2 = 0.9843. F tests of the above regression models indicate that the peak value and its corresponding viewing angle correlate much with surface roughness. The study provides a theoretical base for polarization remote sensing, and impels the rock and city architecture discrimination and minerals mapping.