Comparing two different spectroscopic techniques for the characterization of soil iron oxides: Diffuse versus bi-directional reflectance

Soil spectra, either from bi-directional reflectance (BDRS) and diffuse reflectance spectroscopy (DRS), enable the detection of small amount of hematite and goethite across the VIS-NIR region that provides information regarding the iron oxide minerals content. Good relations between the spectral parameter useful in understanding in iron oxides contents was observed with bi-directional reflectance (BDRS) and diffuse reflectance spectroscopy (DRS) of strongly and moderate weathered soils. Studied soils were Terra Rossa from Mediterranean area derived from karstic dolomitic calcareous environment and a large variety of Oxisols derived from different parent materials ranging from lateritic sediments to basaltic and calcareous rocks from Brazil. The spectra of soil samples were recorded in the laboratory on air-dried-sieved soils. The two different spectra were parameterized by using the amplitude of selected bands in the second derivative of the spectrum of the Kubelka¨CMunk function and by calculating the Munsell color attributes. Hematite and goethite shown very clearly band positions and second-derivative is more sensitive in detecting these two Fe oxides than Munsell color attributes. Significant correlations were found between the spectral and color parameters calculated from the two different spectra. Irrespective of the method, soil goethite and hematite contents were significantly correlated with the amplitudes of the characteristic second-derivative spectral bands at ¡«420 and ¡«535 nm, respectively. The hematite content was correlated with some of the Munsell color parameters, whereas the goethite content was not. Our results suggest that laboratory bi-directional reflectance spectroscopy is as useful as diffuse reflectance spectroscopy to characterize soil Fe oxides. Future work can potentially involve the execution of field-level studies employing a portable spectroradiometer to estimate soil Fe oxides composition. This result supports also the utilization of a sensitive airborne hyperspectral sensor to rapidly and quantitatively evaluate spatial soil information concerning iron oxides content.