Spectropolarimetric characterization of pure and polluted land surfaces

In order to extend remote sensing techniques to reveal further important properties of land surfaces, such as pollution, we have measured and modelled the bidirectional reflectance factor and degree of linear polarization from several particulate surfaces with different materials and different amount of oil contamination. Typically, polarization signals are rather small – of few percents only – and thus precise instruments, robust calibration, and ideal conditions are required to interpret the data. Most strongly, the degree of polarization depends inversely on the brightness of the target. If intensity can not be calibrated, degree of polarization can be used as an indicator of albedo. Oil contamination changes the polarization somewhat, but the effect can be positive or negative, depending on the absorption of oil and host material. Dust cover, air bubbles, and particle surface roughness also affect polarization. However, in real conditions, these and other effects as well as their combinations have not been studied yet, and retrieving unique information typically requires more data and assumptions. As the best observing geometry for polarized remote sensing we suggest a 90° – 120° phase angles range. In certain cases, near-back-scattering angles are indicative and can also be considered. More scattering measurements from different targets, more detailed modelling, and drone tests are recommended for obtaining additional insights in future studies.