Aerosol effects on cloud-precipitation and land-surface processes

Abstract

Aerosols not only directly scatter and absorb solar radiation (denoted as the aerosol direct effects), but also modulate cloud properties by acting as cloud condensation nuclei (CCN) to form cloud droplets (denoted as the aerosol indirect effects). High concentrations of aerosols can overseed cloud droplets to reduce the mean size of cloud droplets, which is hypothesized to increase cloud albedo for the constant cloud liquid water path thereby slowing the warm-rain processes. The first part of this dissertation examines the sensitivity of the aerosol indirect effect to different thermodynamic environments over the tropical ocean. The variability of marine warm cloud properties is derived from satellite multiple sensors, and is normalized by the variability of satellite-derived aerosol index (AI) and reanalysis-derived lower-tropospheric stability (LTS). Global statistics show that increases in AI (polluted air) are associated with reductions in the cloud droplet effective radius (Re), supporting the hypothesis of the aerosol indirect effect. Increase in LTS (strong lower-troposphere stability) is also associated with reductions in the Re, indicating that the stronger inversion prevents dynamical growth of cloud droplets. Marine warm rain processes are estimated from the comparison between cloud-top and column Re, and the global statistics indicate that the warm-rain processes are minimized regardless of the air pollution under a strong temperature inversion, while they are inhibited due to air pollution under unstable thermodynamic conditions. The cloud liquid water path (CLWP) tends to be decreased for higher AI, which does not support the assumption of constant CLW associated with the reduction of Re. Global variability of corrected cloud albedo (CCA: the product of cloud optical depth and cloud fraction) is better explained by the variability of the LTS than by AI. CCA appear to be highest under the strong-inversion regions, and is the first-order property that controls the radiation budget. Local variability of these cloud properties is explained by a combination of AI and LTS better than by either AI or LTS alone. Finally, the spatial mean and the spatial gradient of the aerosol direct and indirect radiative forcing are estimated and compared with the forcing attributed to well-mixed greenhouse gases (GHG) over the tropical ocean. The aerosol direct effect not only reduces global irradiance but also increases diffuse radiation. Diffuse radiation is more homogeneously absorbed by the plant canopy and more efficiently utilized for the plant photosynthesis process than direct radiation. Thus, aerosol loading is expected to increase plant productivity (the aerosol diffuse-radiation effect). The second part of this dissertation examines the spatio-temporal variability of the aerosol diffuse-radiation effect over the eastern U.S., using a sun-shade canopy model. First, satellite and model aerosol optical depth (AOD) products are assimilated via an optimal interpolation technique, and a comparison against the ground-based observations shows that the satellite-model assimilated AOD product is superior to either a satellite or model product. Second, surface albedo, surface radiative temperature, CO2 flux, and sensible/latent heat fluxes in a sun-shade canopy model (Unified Land Model: ULM) are compared with corresponding satellite and ground-based observations. Tuning parameters in ULM are constrained by reducing model-observation discrepancies via the Gauss-Maquardt-Levemberg automatic optimization algorithm. Third, the well-calibrated ULM is run in an off-line model for the warm seasons in 2000 and 2001. Downwelling shortwave radiation is computed with (a control experiment) and without (a potential experiment) assimilated daily AOD in all-sky conditions. The sensitivity experiments (control-potential) show that aerosol loading increases plant productivity in mixed forests and deciduous broadleaf forests in the southeastern U.S., while plant productivity is decreased over the croplands and grasslands. The spatio-temporal variability of aerosol diffuse-radiation effect is well explained by the variability of LAI, cloud optical depth, near-surface atmospheric temperature, and diurnal cycles. Due to the combination of the positive and negative effects, the aerosol diffuse-radiation effect increases plant productivity by only +0.5% in 2001 and -0.09% in 2000 from the potential experiment over the eastern U.S.