Reflectance spectroscopy: Quantitative analysis techniques for remote sensing applications

Abstract

Several methods for the analysis of remotely sensed reflectance data are compared, including empirical methods and scattering theories, both of which are important for solving remote sensing problems. The concept of the photon mean optical path length and the implications for use in modeling reflectance spectra are presented. It is shown that the mean optical path length in a particulate surface is in rough inverse proportion to the square root of the absorption coefficient. Thus, the stronger absorber a material is, the less photons will penetrate into the surface. The concept of apparent absorbance (‐In reflectance) is presented, and it is shown that absorption bands, which are Gaussian in shape when plotted as absorption coefficient (true absorbance) versus photon energy, are also Gaussians in apparent absorbance. However, the Gaussians in apparent absorbance have a smaller intensity and a width which is a factor of √2 larger. An apparent continuum in a reflectance spectrum is modeled as a mathematical function used to isolate a particular absorption feature for analysis. It is shown that a continuum should be removed by dividing it into the reflectance spectrum or subtracting it from the apparent absorbance and that the fitting of Gaussians to absorption features should be done using apparent absorbance versus photon energy. Kubelka‐Munk theory is only valid for materials with small total absorption and for bihemispherical reflectance, which are rarely encountered in geologic remote sensing. It is shown that the recently advocated bidirectional reflectance theories have the potential for use in deriving mineral abundance from a reflectance spectrum.

Does Independent Component Analysis Play a~Role in Unmixing Hyperspectral Data?

“…Negentropy is nonnegative and is equal to zero if and only if has Gaussian distribution. Assuming that components , , are uncorrelated, it follows that (16) which means that finding maximum of Negentropy directions, i.e., maximizing with respect to , is equivalent to minimize the mutual information.…”

Assessing leaf pigment content and activity with a reflectometer

“…Many potential alternatives exist to the simple index approach presented here : these include continuum removal (Clark & Roush, 1984) ; spectral feature-fitting methods (Gao & Goetz, 1994 ; hierarchical foreground \ background analysis (Pinzon et al, 1998) ; derivative spectrometry (Demetriades-Shah, 1990 ;Wessman, 1990) ; and spectral mixture analysis (Adams et al, 1986 ;Wessman et al, 1990 ;Ustin et al, 1993). Stepwise multiple regression has been widely used for analysing reflectance features associated with absorption by biochemical constituents (e.g.…”

Spectral properties of weathered and fresh rock surfaces in the Xiemisitai metallogenic belt, NW Xinjiang, China

“…Moreover, as a complementary tool, the continuum removal technique was used for detecting subtle differences among the spectra [34,35]. Continuum removal is commonly normalized in the range of spectral reflectance from 0 to 1, which is used in spectroscopy for enhancement of the intensity of the characteristic peaks and absorption features [34,36]. Therefore, we used continuum removal to normalize the spectral data by reducing the variation in individual spectra (Figures 3-8b), with the goal of overcoming differing backgrounds and interference from spectrally overlapping minerals.…”