Observational evidence for the convective transport of dust over the Central United States

Abstract Bulk aerosol composition and aerosol size distributions measured aboard the DC‐8 aircraft during the Deep Convective Clouds and Chemistry Experiment mission in May/June 2012 were used to investigate the transport of mineral dust through nine storms encountered over Colorado and Oklahoma. Measurements made at low altitudes (<5 km mean sea level (MSL)) in the storm inflow region were compared to those made in cirrus anvils (altitude > 9 km MSL). Storm mean outflow Ca 2+ mass concentrations and total coarse (1 µm < diameter < 5 µm) aerosol volume ( V c ) were comparable to mean inflow values as demonstrated by average outflow/inflow ratios greater than 0.5. A positive relationship between Ca 2+ , V c , ice water content, and large (diameter > 50 µm) ice particle number concentrations was not evident; thus, the influence of ice shatter on these measurements was assumed small. Mean inflow aerosol number concentrations calculated over a diameter range (0.5 µm < diameter < 5.0 µm) relevant for proxy ice nuclei ( N PIN ) were ~15–300 times higher than ice particle concentrations for all storms. Ratios of predicted interstitial N PIN (calculated as the difference between inflow N PIN and ice particle concentrations) and inflow N PIN were consistent with those calculated for Ca 2+ and V c and indicated that on average less than 10% of the ingested N PIN were activated as ice nuclei during anvil formation. Deep convection may therefore represent an efficient transport mechanism for dust to the upper troposphere where these particles can function as ice nuclei cirrus forming in situ.