Seasonal Variability of Aerosol-Cloud Interaction estimated from Long-term Satellite Analysis

The present study investigated the correlations between aerosol and cloud parameters derived from satellite remote sensing to estimate properties of aerosol-cloud interactions. The improvements in satellite instrumentations and retrieval methods have provided valuable data to minimize uncertainties in assessments of aerosol-cloud interactions. The retrievals for aerosols and cloud parameters are not coincident, however, we use spatial and temporal averaged data, which are considered to have typical properties in each region and period. In this study, better evaluations are achieved to use monthly data from 2001 to 2010 at once, because seasonal variations are smaller than annual ones. The global statistics showed that effective particle radius and optical thickness of low clouds correlate well with column number concentration of the aerosol particles in small – moderate amount of atmospheric aerosol loading (about Na < 10⁹ [particles/cm²]), which are consistent with an aerosol indirect effect. The steeper regression slopes and better correlation coefficients are shown in case of small liquid water path (LWP) condition than large one. To study the seasonal variability and regional properties, we divided the area covered by satellite observations and analyzed correlations between aerosol and cloud parameters. In mid-latitude, we find regression slopes between aerosol number concentrations and cloud parameters are steeper, it means, aerosol affects clouds greater in winter than summer, and in northern than southern hemisphere. That seasonal variation does not appear in tropics and the slope of regression line is almost same in mid-latitude summer. It is considered that cloud formation process is strongly affected by atmospheric dynamical processes in tropics and mid-latitude summer, that aerosols' effect is appeared relatively weak. In case of turbid atmosphere, inverse trends between aerosol and cloud microphysics parameters are appeared. It is found that cloud effective radius is positively and cloud optical thickness is negatively correlated with column number concentration of aerosol in large amount of atmospheric aerosol loading (about Na > 10⁹ [particles/cm²]). These inverse trends are rarely or weakly appeared in the southern subtropics. The positive trend of cloud effective radius is not found in case of large LWP conditions. We will discuss regional characteristics of these trends and differences derived from aerosol compositions. Figure: Correlation plots between column aerosol number and (a) cloud effective radius, (b) cloud optical thickness using 0.25º grids daily averaged MODIS data for 10 years in northern mid-latitude (35 – 45 Nº).