The field of vegetation remote sensing has historically been dominated by radiative transfer models (RTMs) employing spectral reflectance as a means of characterising the interaction between vegetation and incident radiation. In recent years, polarimetry has received increasing attention as a complementary source of information in remote sensing, driven in part by advances in polarimetric instrumentation and the increasing availability of measurements capable of resolving the polarisation state of reflected radiation. The polarisation state of light contains information on the optical and structural properties of vegetation that is not captured by conventional radiometric measurements.

This study explores whether polarimetric observations of leaves contain relationships between illumination geometry, viewing geometry, and leaf degree and angle of linear polarisation (DoLP, AoLP), with the broader aim of assessing if polarimetry could enhance vegetation remote sensing by revealing information on physical and environmental vegetation characteristics. Two radiative transfer models, PROPOLAR and DART, are assessed in their ability to reproduce vegetation polarisation. The modelling component further aims to assess whether current RTM approaches can provide a sufficiently accurate physical description of vegetation polarisation to support the future development of a retrieval method between polarimetric observations and vegetation physical and environmental traits.

The results show that both DART and PROPOLAR, in their current implementations, are unable to adequately reproduce the laboratory measurements, with discrepancies potentially arising from their respective implementations of leaf microfacet properties. The laboratory measurements nevertheless show the existence of systematic relationships between observation geometry, leaf ar
... Read more