Quantifying the Specular-to-Diffuse Transition: Spectral HDR and BRDF Mapping of Surface Morphology Effects for SDA Modeling

Moses Kodur, Surface Optics Corp; Samuel Dummer, Surface Optics Corp
https://doi.org/10.64861/GVUO9955

Keywords: Roughness Crossover, Diffuse Gold, Satellite Characterization, SDA, BRDF, HDR Reflectance, Specular vs Diffuse, MWIR/LWIR, Glint, EO/IR Signature Modeling, Surface Metrology, Specular Fraction, Emissivity, Hemispherical Emittance

Abstract:

Electro-optical and infrared signature models used for Space Domain Awareness (SDA) commonly treat surface reflectance as a static material constant and classify surfaces as either specular or diffuse. This work presents an experimental dataset demonstrating that the partitioning of reflected energy between specular and diffuse channels is instead a strong function of the ratio between wavelength and surface roughness, and that this partitioning can shift by more than an order of magnitude while total hemispherical reflectance changes comparatively little. A gradient of four commercially available gold-coated ground-glass diffusers spanning nominal 120, 220, 600, and 1500 grit polishes was characterized using three instruments: integrating-sphere directional-hemispherical reflectance (DHR) from 250 to 2500 nm, hemiellipsoidal hemispherical-directional reflectance (HDR) with mechanical specular exclusion from 2 to 50 µm, and bidirectional reflectance distribution function (BRDF) measurement in discrete mid-wave and longwave infrared bands. Specular fraction rises from below 0.02 in the visible to above 0.98 in the long-wave infrared for the finest polish, with crossover wavelength scaling with effective RMS roughness as predicted by scalar diffraction theory. Two measurement results are emphasized. First, the reported specular fraction depends on how an instrument defines the specular channel: the integrating-sphere and hemiellipsoidal instruments report different nonzero floors on diffuse-dominated surfaces. Second, BRDF mapping confirms that the floor corresponds to real near-specular concentration rather than a Lambertian background; for the 120-grit sample at 20° incidence, integration over 6°-7° cones bracketing the SOC-100’s 6.7° blocker contains approximately twice the Lambertian fraction. Quantitative equality between instruments is not claimed because the measurements differ in illumination and collection geometry, spectral weighting, sampled area, polarization treatment, and reference standard. Because the samples are catalog parts available under fixed part numbers, the dataset is offered as a candidate reference artifact for inter-laboratory comparison and for validation of parametric scattering models used in signature prediction.

Date of Conference: September 15-18, 2026
Track: Satellite Characterization

 

View Paper