Contrasting the impact of Saharan Dust on ice cloud properties and rainfall over West Africa and the Atlantic Ocean during summer

From observations of the Infrared Atmospheric Sounding Interferometer (IASI) the atmospheric loading of Saharan dust as well as the properties of ice clouds (cirrus and deep convection) are determined. These are then be used to infer the relationships between dust loading and ice cloud properties such as optical depth, effective diameter and ice water path. From these infrared observations a statistical analysis of moderately thick ice clouds such as cirrus or glaciated altocumulus is performed, also addressing the question of dust initiating early glaciation in moderately thick clouds. In order to account for different meteorological environments, cloud top temperature is used to statistically constrain dust-affected observations to background distributions of cloud properties by Bayesian methods. Therefore the background (low dust loading) probability density distributions of (ice) cloud optical depth, effective diameter and ice water path as well as rainfall (from the Tropical Rainfall Measurement Mission TRMM) are calculated as functions of cloud top temperature intervals. The observation of these parameters under higher dust loading (with potentially deviating cloud top temperature distribution) thus is projected onto the cloud top temperature distribution of the background cloud sample and the difference in the probability density distribution is estimated through histogram deviations. Cloud property distributions are evaluated for background conditions (i.e. Dust AOD_0.55 < 0.2), moderately dusty environments (0.2 < Dust AOD_0.55 < 0.75) and high dust loadings (Dust AOD_0.55 > 0.75) for contrasting these different conditions in a statistical way. Relationships between dust and ice clouds are compared for two different regions: the West African Sahel, where the northward progression of the West African Monsoon during summer frequently generates strong mesoscale convective cloud clusters (also themselves impacting on dust emission) close to dust source regions, and the tropical North Atlantic Ocean, where summertime dust outbreaks are frequent features and mesoscale convection occasionally develops into tropical depressions or storms. Two years of summertime observations have been used for the statistical samples. The analysis reveals an overall insignificant average change in ice optical depth under dusty conditions for the clouds observed. However, the distribution of ice cloud optical depth with cloud top temperature is subject to significant change with different magnitudes over land and ocean. An increase in ice optical depth for the cloud top temperature range 255K-265K indicates an early freezing effect in moderately cold clouds (e.g. altocumulus). The strongest dust related effects are observed in cloud top ice crystal effective diameter, and here also a strong contrast between land and ocean is found. Over the Atlantic Ocean the effective diameter average is slightly reduced under influence of high dust loadings, while the average reduction is much more pronounced over land, where mean deviations are -1.1 micrometer and -2.6 micrometer for moderate and high dust loadings respectively. The results clearly indicate that the dust influence reduces the bimodality of the effective diameter distributions, thus the magnitude of the skewness is significantly reduced. Analysis of TRMM rain rates together with IASI dust observations shows a consistent significant reduction of overall rainfall with increased dust over both, the Atlantic Ocean and West Africa. Also variance and skewness of the rainfall distribution change with increasing dust loading towards more light rain and strongly reduced moderate rain rates (1-3mm/hr), while strong precipitation remains almost unaffected by dust presence.