Conversion from Ross-Li to Rahman-Pinty-Verstraete model parameterizations of surface bidirectional reflectance

Accurate modeling of anisotropic bidirectional surface reflectance is essential for multi-angle aerosol retrievals over land, where surface contributions often dominate the top-of-atmosphere signal. Two widely used semi-empirical bidirectional reflectance factor (BRF) formulations are the Ross-Li model, adopted by the MODIS/MAIAC land surface product, and the Rahman-Pinty-Verstraete (RPV) model, used operationally in MISR and the forthcoming MAIA aerosol retrieval algorithms. Differences in surface parameterization across retrieval systems complicate the transfer and reuse of surface reflectance constraints. To address this issue, we develop a computationally efficient algorithm that converts Ross-Li BRF parameters into their RPV counterparts. The method exploits the quasi-linearity of the RPV model in logarithmic space to enable initialization-free, linear least-squares estimation of spectral weight and angular-shape parameters, optionally imposing spectral invariance constraints. Additional steps convert the exponential scattering term of the modified RPV into the Henyey-Greenstein phase function of RPV and ensure physical consistency by conserving directional hemispherical reflectance through analytic hotspot adjustment. Using MAIAC BRF parameters as input, we evaluate the conversion for urban, desert, and vegetated surfaces. The resulting RPV BRFs generally reproduce the spectral magnitude and angular structure of the Ross-Li simulations, with relatively larger differences confined to the hotspot region and extreme viewing geometries. Based on this conversion, we assess the effect of the two surface models by performing retrievals using combined MISR and MODIS observations. Comparison against the AEROENT reference data shows that the two retrievals exhibit comparable overall performance. The RPV-based retrieval yields slightly improved agreement in aerosol optical depth, whereas the Ross–Li-based retrieval provides more consistent estimates of aerosol single-scattering albedo and particle size.