Impact of Saharan dust on North Atlantic marine stratocumulus clouds: importance of the semidirect effect

Abstract

Abstract. One component of aerosol–cloud interactions (ACI) involves dust and marine stratocumulus clouds (MSc). Few observational studies have focused on dust–MSc interactions, and thus this effect remains poorly quantified. We use observations from multiple sensors in the NASA A-Train satellite constellation from 2004 to 2012 to obtain estimates of the aerosol–cloud radiative effect, including its uncertainty, of dust aerosol influencing Atlantic MSc off the coast of northern Africa between 45° W and 15° E and between 0 and 35° N. To calculate the aerosol–cloud radiative effect, we use two methods following Quaas et al. (2008) (Method 1) and Chen et al. (2014) (Method 2). These two methods yield similar results of −1.5 ± 1.4 and −1.5 ± 1.6 W m−2, respectively, for the annual mean aerosol–cloud radiative effect. Thus, Saharan dust modifies MSc in a way that acts to cool the planet. There is a strong seasonal variation, with the aerosol–cloud radiative effect switching from significantly negative during the boreal summer to weakly positive during boreal winter. Method 1 (Method 2) yields −3.8 ± 2.5 (−4.3 ± 4.1) during summer and 1 ± 2.9 (0.6 ± 1) W m−2 during winter. In Method 1, the aerosol–cloud radiative effect can be decomposed into two terms, one representing the first aerosol indirect effect and the second representing the combination of the second aerosol indirect effect and the semidirect effect (i.e., changes in liquid water path and cloud fraction in response to changes in absorbing aerosols and local heating). The first aerosol indirect effect is relatively small, varying from −0.7 ± 0.6 in summer to 0.1 ± 0.5 W m−2 in winter. The second term, however, dominates the overall radiative effect, varying from −3.2 ± 2.5 in summer to 0.9 ± 2.9 W m−2 during winter. Studies show that the semidirect effect can result in a negative (i.e., absorbing aerosol lies above low clouds like MSc) or positive (i.e., absorbing aerosol lies within low clouds) aerosol–cloud radiative effect. The semipermanent MSc are low and confined within the boundary layer. CALIPSO shows that 61.8 ± 12.6 % of Saharan dust resides above North Atlantic MSc during summer for our study area. This is consistent with a relatively weak first aerosol indirect effect and also suggests the second aerosol indirect effect plus semidirect effect (the second term in Method 1) is dominated by the semidirect effect. In contrast, the percentage of Saharan dust above North Atlantic MSc in winter is 11.9 ± 10.9 %, which is much lower than in summer. CALIPSO also shows that 88.3 ± 8.5 % of dust resides below 2.2 km the winter average of MSc top height. During summer, however, there are two peaks, with 35.6 ± 13 % below 1.9 km (summer average of MSc top height) and 44.4 ± 9.2 % between 2 and 4 km. Because the aerosol–cloud radiative effect is positive during winter, and is also dominated by the second term, this again supports the importance of the semidirect effect. We conclude that Saharan dust–MSc interactions off the coast of northern Africa are likely dominated by the semidirect effect.