Physical optics approach

The term physical optics is generally used to refer to optical effects that cannot be explained based on simplified geometrical optics. Alternatively, perhaps more appropriately, the term wave optics is used, as this refers to the nature of light as an electromagnetic wave. The most fundamental approach to optics would require solving Maxwell’s equations with high spatial and temporal resolution. Since this level is usually not achievable for practical applications, a significant degree of abstraction and simplification is necessary. The methods applied in geometrical optics are, in fact, a result of farreaching abstraction. The present work attempts to improve the level of abstraction by enriching geometrical optics with the most relevant models of physical optics in a comprehensible and practically relevant way. The more fundamental level of physical optics allows more accurate models that potentially require significantly fewer input data than empirical models. Beyond that, more fundamentally based models are also more likely to meet future, yet unknown, requirements.
The field of physical optics naturally covers a wide range of effects ranging from non-linear optics at high field strengths in laser physics, over interferometry or Doppler-shifts in astronomy to nanooptics in material science. The present work identifies the processes with the most significant impact on solar-related energy flows. For this purpose, suitable models are adapted for implementation. Explicitly, these are:

• consideration of global radiation spectrum and spectral dependencies
• consideration of polarisation state
• consideration of fundamental material properties (refractive indices)
• consideration of angular and spectral dependencies
• consideration of surface roughness and subsurface scattering
• consideration of coherence effects on thin-flims (coatings)
• consideration of attenuation in bulk materials

The reason to follow this fundamental approach is two-fold. First, only physically based models are able to capture a satisfyingly wide range of practically relevant optical phenomena. Due to the intrinsically complex nature of light-matter interactions, simple models are only valid within strictly limited boundaries. Consequently, there is a constant need to apply additional ad-hoc models or workarounds to cover effects beyond the scope of the simplified model (e.g. wavelength dependences, polarisations effects etc.). Second, physical models require significantly fewer input parameters to achieve the same level of accuracy as empirical, data-driven models, as essential information on the processes is contained within the models in the form of fundamental relations. Therefore even limited inputs or constant approximations can potentially provide spectrally resolved results with higher accuracy than the models used in international standards.