| Summary: | Atmospheric aerosols modulate the radiative budget and ambient air quality of the atmosphere, thus, there is a need to develop both analytical and computational methodological techniques that determine their physical, chemical and optical properties in order to characterize and model their effects. This thesis embodies the results of the derivation of radiative characteristics of the atmosphere over Nairobi (2006-2008), Mbita (2007) and Malindi (2008) using aerosol data obtained from sun spectrophotometry. Aerosol optical depths (τ), single scattering albedo (ω), angstrom exponent (α), asymmetry factor ( ), real ( ) and imaginary ( ) refractive indices at zero Solar Zenith Angle (SZA) were derived through AErosol RObotic NETwork (AERONET) framework. Temporal and spatial characteristics in τ and α were investigated using multivariate techniques viz. Principal Component Analysis (PCA), Partial Least Squares Discriminant Analysis (PLS-DA), Principal Component Regression (PCR) and Hierarchical Cluster Analysis (HCA). Annual averages of the optical properties together with selected physico-chemical properties i.e. aerosol number densities and extinction cross section were determined. The Coupled Ocean and Atmosphere Radiative Transfer (COART) model was used to solve the radiative transfer equation (RTE) for an atmosphere assumed to be purely impacted by aerosols of different sizes and estimated their radiative impacts. The use of multivariate chemometric techniques revealed that temporal and spatial characteristics of both τ and α over the study sites are modulated by weekly total rainfall, relative humidity, temperature, aerosol hygroscopic properties, aerosol burden, aerosol mode of generation and composition, both local air circulation and urban heat island effects. There was no significant spectral dependence in ω, and both and at zero SZA over the study sites. Comparison of the measured τ and α from AERONET at λ = 500 nm and 440/675 nm respectively was achieved by utilizing Moderate Resolution Imaging ...
|
|---|