AAAR 37th Annual Conference October 14 - October 18, 2019 Oregon Convention Center Portland, Oregon, USA
Abstract View
Estimating Uncertainties in Refractive Index Retrievals from Optical Closure Calculations using Full Aerosol Size Distributions
ALEXANDER FRIE, Roya Bahreini, University of California, Riverside
Abstract Number: 216 Working Group: Aerosol Physics
Abstract The refractive index (m) is the fundamental descriptor of a material’s optical properties. For aerosols, m can be used to calculate their direct radiative effect within the atmosphere. Spherical aerosol refractive indices are commonly calculated using different optical closure methods, where the observed optical properties are compared to those predicted using Mie theory for different m values. One method uses a series of monodisperse aerosol populations and associated optical measurements to calculate m (Size Selected), and another uses the bulk size distribution and associated optical measurements (Full Distribution). Due to the time needed to switch between multiple diameters, the Size Selected method has limited field and laboratory applications if aerosol properties are highly dynamic. Alternatively, Full Distribution measurements are faster and have potential field applications. Given this, it is important to understand the uncertainties associated with the Full Distribution retrievals. Due to the complexity of Mie theory, and its computational expense, propagation of measurement uncertainties through the m retrievals has been difficult. Despite this, constraining the uncertainty in m retrievals is essential as it may change depending on the m value¸ observational uncertainties, and the aerosol size distribution. Here a pseudo-Monte Carlo tool is developed which propagates instrumental uncertainties through the m calculations via simulating multiple possible “true” m values, perturbing size distribution and optical characteristics with instrumental uncertainties, and recalculating m using the perturbed values. Repeating these calculations a large number of times yields a probability distribution of m values which could have yielded the originally retrieved m, indicating its uncertainty range. Results from this tool are demonstrated using seven representative aerosol size distributions over a series of m values and a series of uncertainty conditions. These simulations reveal the importance of considering aerosol size distribution, m, and instrumental uncertainties when reporting aerosol m values.