ISCCP at 30

Influence of aerosols on mesoscale convective systems inferred from ISCCP and A-Train datasets Rong Fu & Sudip Chakraborty Jackson School of Geosciences, The University of Texas at Austin ISCCP at 30, Grove School of Engineering, Steinman Hall, City College of New York, NY, NY April 23 rd, 2013 Guillermo Rein

Stevens and Feingold, 2009: Despite decades of research, it has proved frustratingly difficult to establish climatically meaningful relationships among the aerosol, clouds and precipitation. As a result, the climatic effect of the aerosol remains controversial. The absence of observations of cloud life cycles that show precipitation reducing cloud lifetimes underscores how loosely the term ‘lifetime’ has come to be applied. Motivation:

Rosenfeld et al. (2008): Their hypothesis: Aerosols suppress rainfall in shallow convection, but invigorate rainfall in deep convection. Implication of their hypothesis: aerosols can intensify both drought by suppressing light rain and heavy rainfall (flood) Main uncertainty: a clear relationship between intensification of deep convection and aerosol concentration has not been established observationally (e.g., Williams 2002).

What observational relationships should we look for? Excepted relationship with increase of aerosols loading based on this hypothesis: Enhanced ice vs. liquid ratio Increased ice water content in convective anvils Increased convective life time. Hypothesis of Rosenfeld et al., 2008 Delaying precipitation during the growing and mature stages Allow more glaciation of cloud droplets Stronger updraft More small ice particles in the UT Longer convective lifetime Clean Polluted

Challenge: How can we put A-Train measurements in the context of evolution or life cycle of the mesoscale convective systems?

Datasets: SatellitesPurpose Aura MLS CO Data.To detect deep convection and pollution transport. CALIPSO Vertical feature MaskTo detect vertical distribution of aerosols (use only samples with |CAD| score>70). CloudSat Radar, Cloud Water Content To locate deep convection using CWC and CWP. MODISAerosols optical depth (AOD) and cloud effective radii (ER) ISCCP Convective trackingTo track the deep convection and life cycle of convective systems. Overlap period among these datasets: June 2006-June 2008

Domains and periods of our analysis: South America: 5N-15S, 40-80W, Congo: 10N-10S, 10W-40E for the period of June 2006-May 2008, SE Asia: 0-40N, E, for the periods of June-August, 2006, 2007, June cases mesoscale convective systems with life-time > 6 hours & d>100Km were located with A-train measurements, 355 cases in growing phase, 401 in mature phase and 210 during decay phase. Aerosols types: SE Asia Congo Amazon Smokedust Polluted dust

Linking A-Train measurements to the life cycle of convection as indicated by the ISCCP Jan 21-22, 2007

Increases of cloud ice/liquid water ratio are significantly correlated with increases of number of polluted pixels (AOD>0.3). This relationship is consistent with the key mechanism that enable aerosols to invigorate convective systems proposed by Rosenfeld et al. (2008). GrowingMature Decay Clean polluted Clean polluted More ice More liquid MODIS CloudSat

Does IWC in convective anvils increase with aerosol loading? Ice water content in convective anvils appear to increase with ambient aerosol AOD in S. Asia and Congo during active phase of the convective systems. Cannot detect change of IWC of convective anvils with ambient aerosols over S. America. Number of pixels with AOD>0.3 (MODIS) Ambient RH at 850 hPa (MERRA) Aura MLS IWC at 215 hPa (mg/m 3 ) Active phasemature phase

The variance of the cloud life time appears to be mainly linked to the initial cloud dynamic properties (size, convective fraction and number of convective cores). Initial ambient aerosols loading may explain a small fraction of the cloud- life time variation. Could aerosols influence life-time of the mesoscale convective systems? Stepwise regression

Can we detect convective transport of aerosols? Smoke Calipso L3 aerosol profile data, Huang et al. 2013

Aerosol layers are mostly detected at the height of convective detrainment layer HDL follow the method stated in Mullendore et al in 2009 CloudSat Detrainment layer Calipso Cloud Top height Calipso Height of Aerosol Layer

Vertical distribution of Aerosol layers at different stages of convective life cycle. More aerosols layer are detected in the vicinity of growing convections than in vicinity of decaying convection. growingmaturedecay

Relative importance of convective dynamic properties and wind shear to convective transport of aerosols Stepwise Regression

Preliminary conclusions: A-train measurements + ISCCP cloud tracking data show good potential for detecting climate relevant relationships between cloud structure and aerosols at various stages of convective life time. An increase of cloud ice/liquid water ratio in mesoscale convective system is significantly correlated with an increase of ambient aerosol loading, supporting one aspect of the hypothesis that aerosols invigorate atmospheric deep convection. – S. Asia: IWC in convective anvil and convective life-time also appear to increase with ambient aerosol loading – Congo: IWC in convective anvil increase with aerosols loading – Amazon: IWC and convective life-time do not appear to link to ambient aerosol loading Convective transport of aerosols to the upper troposphere appear to be most influenced by number of convective cores, size and wind shear during the active phase of the mesoscale convective systems.