Spectral Signatures in Shortwave Radiation Measurements to Derive Cloud and Aerosol Properties

Abstract

The amplitude and spectral shape of shortwave radiation are used to retrieve aerosol and cloud properties from airborne and ground based measurements. By interacting with clouds and aerosols in the Earth’s atmosphere, the wavelength-dependent radiation emitted by the sun is modified. This thesis presents the change in radiation due to absorption and scattering by clouds and aerosols, which result in distinct spectral signatures in shortwave radiation spectra. The spectral signature in shortwave radiation due to aerosols is quantified by airborne measurements of irradiance above and below aerosol layers. This radiative effect is quantified by the relative forcing efficiency, which is used to compare the impact of aerosols from different air masses, locations, and time of day. The relative forcing efficiency is the net irradiance change due to the presence of aerosols normalized by aerosol optical thickness and incident irradiance. It is shown to vary by less than 20% per unit of midvisible aerosol optical thickness for aerosols sampled during 4 different experiments, except for highly absorbing aerosols near Mexico City. The similarity in relative forcing efficiency for these experiments, not expected a priori, suggests that this quantity is constrained for various types of aerosols with differing scattering and absorption characteristics even when surface albedo differs. To estimate the radiative effect of aerosols sampled in the Los Angeles basin during one of the experiments, where no concurrent measurements of optical thickness with spectral irradiance were available, a new iterative technique was devised to use aerosol optical thickness measurements from another airborne platform. Cloud-transmitted zenith radiance spectra were measured from the ground in Boulder, Colorado. In these measurements, spectral signatures of cloud optical and microphysical properties were uncovered. The spectral signatures are the result of radiation that is transmitted through clouds, where ice or liquid water cloud particles modulate the radiation by absorbing and scattering incident light in a wavelength-dependent manner. Typically, the magnitudes of radiance at 2 wavelengths have been used to retrieve cloud properties, but by using wavelength-dependent features more sensitivity to cloud microphysical properties is obtained. This thesis presents a method to analyze wavelength-dependent signal, where spectral features such as slopes, curvatures, and shifts in locations of maxima and minima are parameterized. These spectral features found in normalized radiance are quantified by introducing 15 parameters. These 15 parameters form the basis of a new generalized retrieval obtaining cloud optical thickness ($au$), effective radius (re), and ther- modynamic phase ($phi$). When applied to a liquid water cloud case, this retrieval matched a measured transmittance spectrum with a smaller root mean square difference over the entire spectrum (3.1%) than two other methods (up to 6.4%). To quantify the retrieval over all possible combinations of $au$, re, and $phi$, simulated measurements were used in conjunction with realistic measurement and model error characteristics. By combining these error characteristics within the GEneralized Nonlinear Retrieval Analysis (GENRA) a solution probability distributions can be built. The information of cloud properties contained within cloud-transmitted radiance is greater on average for liquid water clouds than for ice clouds. For all possible combinations of cloud properties, radiance transmitted through clouds with $au$ { extless}20 contain the most information on cloud properties, indicating that the 15 parameters have greatest sensitivity to cloud properties of optically thin clouds ($au$ { extless}20). Of the 15 parameters, only 10 are required to retrieve accurately $au$, re, and $phi$ for any cloud except for ice clouds with $au$ { extgreater}25 and re { extgreater}30 µm. Using this retrieval, the correct thermodynamic phase is determined from transmittance with a probability greater than 99.4% for horizontally homogeneous clouds that contain either ice or liquid water cloud particles.

Samuel LeBlanc

Research Scientist

Atmospheric Scientist studying colors of clouds and aerosol to better measure and understand climate