An Empirical Function for Bidirectional Reflectance Characterization for Smoke Aerosols Using Multi-angular Airborne Measurements

Abstract:

Using airborne Cloud Absorption Radiometer (CAR) reflectance measurements of smoke, an empirical relationship between reflectances measured at different sun-satellite geometry is established, in this study. It is observed that reflectance of smoke aerosol at any viewing zenith angle can be computed using a linear combination of reflectance at two viewing zenith angles. One of them should be less than 30° and other must be greater than 60°. We found that the parameters of the linear combination computation follow a third order polynomial function of the viewing geometry.

Similar relationships were also established for different relative azimuth angles. Reflectance at any azimuth angle can be written as a linear combination of measurements at two different azimuth angles. One must be in the forward scattering direction and the other in backward scattering, with both close to the principal plane. These relationships allowed us to create an Angular Distribution Model (ADM) for smoke, which can estimate reflectances in any direction based on measurements taken in four view directions.

The model was tested by calculating the ADM parameters using CAR data from the SCAR-B campaign, and applying these parameters to different smoke cases at three spectral channels (340nm, 380nm and 470nm). We also tested our modelled smoke ADM formulas with Absorbing Aerosol Index (AAI) directly computed from the CAR data, based on 340nm and 380nm, which is probably the first study to analyze the complete multi-angular distribution of AAI for smoke aerosols. The RMSE (and mean error) of predicted reflectance for SCAR-B and ARCTAS smoke ADMs were found to be 0.002 (1.5%) and 0.047 (6%), respectively. The accuracy of the ADM formulation is also tested through radiative transfer simulations for a wide variety of situations (varying smoke loading, underlying surface types, etc.).