|
High clouds overlapping low clouds
We show that low cloud fraction has a strong land-ocean contrast with oceanic values double those over land. Major low cloud regimes include not only the eastern ocean boundary stratocumulus and shallow cumulus but also those associated with cold air outbreaks downwind of wintertime continents and land stratus over particular geographic areas. Globally, about 30% of low clouds are overlapped by high clouds. The overlap rate exhibits strong spatial variability ranging from higher than 90% in the tropics to less than 5% in subsidence areas, and is anti-correlated with subsidence rate and low cloud fraction. The zonal mean of vertical separation between cloud layers is never smaller than 5 km and its zonal variation closely follows that of tropopause height, implying a tight connection with tropopause dynamics. Possible impacts of cloud overlap on low clouds are discussed.
For details please see the manuscript .
Human-induced lightning and its impact on ozone
We show that lightning in convective storms is very sensitive to aerosol concentrations. Over our study region, an increase of 0.1 of aerosol optical depth results can double or quadruple lightning flash rate. We demonstrate that this connection is not due to other factors than aerosols themselves. Such aerosol related increase in lightning substantially enhances production of ozone production by increasing the concentration of NOx.
In light of the anthropogenic increase of aerosols, we propose a non-linear relationship between aerosol concentration and lightning flash rate. Once we take this relationship into account and parameterize it in a global model we are able to show that as a result of human-induced lightning change by aerosols, tropospheric ozone has increased significantly, especially in the upper troposphere within the tropics.
For details please see the published paper.
Intercontinental aerosol transport
Yu et al. (2012) examines the contributions of international source versus domestic production of aerosols over the North America using 3-D aerosol satellite observation as well as chemical transport model results. Unexpectedly, foreign aerosols, after surviving thousands of miles of travel, contribute almost the same amount of aerosol mass as the total domestic production. Moreover, dust dominate the foreing source, not pollution particles. Dust arrive mostly at high altitudes (see figure) and most of the pollution does not survive the long-transport. The imported aerosols have significant impacts on the energy balance of regional climate system. They can also affect the regional climate through interacting with clouds and precipitation.
For details please see the published paper.
aerosol-convection-lightning connections
In this study we present clear evidence of aerosols invigorating convection and enchancing lightning activity of maritime convetive clouds. We are able to effectively rule out important possibility of co-varying meteorological factors. We present specific observationl signatures of the aerosol-induced changes using satellite-based radar and MODIS data. The left figure shows the spread of aerosol precusor gas and aerosol (upper panels) and the correlation between aerosol concentration and lightning activity (lower panels).
For details please see the published paper.
We use data from the NASA's A-Train to analyze the trade cumulus clouds' response to influx of aerosols in a natural experiment. Our study area covers millions of square kilometers and the study period lasts for three months. Our results show increased aerosol concentration decreases droplet size, suppresses warm precipitation, increases cloud optical depth and increases cloud amount. We were able to effectively demonstrate that these efffects are only result of aerosol changes. It represents one of the first observational evidences for the so-called aerosol cloud amount effect (or lifetime effect) on large spatial and long temperal scales. The left figure shows the aerosol plume (A) and plumes of increased cloud optical depth, decreased droplet size and increased cloud amount (B,C,D, respectively).
For detailed discussion see published paper.
Cloud fileds have tramendous variabilities and often look chaotic and highly heterogeneous. An example of a cloud field is shown in panel A of the figure to the left. In this study it is shown that these clouds have an emerging behavior in terms of their sizes. The size distribution of these clouds follows nicely with a power-law. More importantly, cloud organization shows self-organizing characteristics. A statistical mechanics approach is suggested to study cloud macro-organization. An example is shown where based on two simple rules cloud organization can be simulated, suggesting the order of cloud behavior has its root in random interactions.
For detailed discussion see published paper.
Estimating glaciation temperature
The latent heat release by deep convective clouds is the engine for atmosphere's general circulation. The vertical structure of latent heat release depends on many factors and one of them is the phase transition of cloud hydrometeors. Knowledge about this structure is critical for acurately estimating latent heat release. We propose a technique to retrieve cloud glaciation temperature, the temperature when all cloud condensate are in ice phase, from passive satellite sensors. Its viability is demonstrated both in a physical conceptual model and by air-borne measurements. The figure on the top shows the conceptual model of cloud hydrometer size growth in the vertical. The panel on the bottom is showing an observation by airborne instrument and retrievals based on the measurements.
For details of the technique and its important applications see paper in GRL
Deep convective cloud statistics
In this study we present statistics of both macro- and micro- physical properties from MODIS measurements of deep convective clouds. With MODIS radiance and cloud retrievals we show some interesting relationships and invariant behaviors. An important relationship is shown between the size of cloud ice particles and cloud brightness temperature. They are positively correlated suggesting size sorting. This relationship is universally obeyed although it does have regional characteristics. This relationship has particularly important implications for observational study of aerosol effect on ice particle sizes. The invariant behavior is demonstrated by the left figure. The cloud optical depth distribution as well as that of brightness temperature does not have significant inter-annual variations for a fixed location despite interannual variations in meteorology. This has important implications for high cloud physics and feedbacks.
For details discussions on these observations please see paper at J. Climate.
Complicated interactions between aerosols and deep convective clouds have given rise many studies that sometimes show different and even seemingly contradictory results. In this study we use model simulations and show that vertical wind-shear is a dominant factor in qualitatively determining whether aerosols invigorate or suppress convection and precipitation. Specifically, aerosols always suppress convection when wind-shear is strong while they can either suppress or invigorate convection otherwise. This provides a first-order regime separator for understanding aerosol-deep convective cloud interactions. The left figure shows cloud responses to aerosol perturbation under different conditions of wind shear and humidity.
Detailed discussion and results can be found in our JGR paper.
Caveats of passive remote sensing
Using passive sensor data to do aerosol-cloud interaction is faced with many caveats and difficulties. In this study we confront these difficulties and use an approach that relates cloud properties and aerosol properties within a mesoscale sized domain. What we found was that aerosol affects clouds differently when the environmental conditions are different. We also clearly proposed many potential dangers of using correlations between aerosol and cloud retrievals from passive satellite sensors and addressed them in our study. The left figure shows how aerosols change cloud droplet size in different ways when the environmental conditions are different. We show with cloud resolving model simulations that observed behavior can be indeed explained using different aerosol and meteorological conditions.
Details on these results are in this paer.
| Home | Research | Publications | About me | News and Links |