Impacts Of Dust Aerosol On The Vertical Structures Of Clouds And Precipitation

Clouds and precipitation play critical roles not only in the Earth's energy balance and water cycle,but also in climate changes.Clouds and precipitation could undergo significant changes due to the increase in atmospheric aerosols.Mineral dust is one of the most abundant and widely distributed aerosols in the troposphere.Dust aerosol could act as cloud condensation nuclei(CCN)and ice nuclei(IN)to modify the microphysical properties of clouds and precipitation,thereby affecting the climate system.However,there are large uncertainties in the representation of aerosol-cloud-precipitation interactions(ACPI)in current climate projections,according to the latest report by the Intergovernmental Panel on Climate Change(IPCC).Therefore,more efforts should be made to investigate the dust aerosol interactions with clouds and precipitation.This study aims to reduce the uncertainties in dust aerosol interactions with clouds and precipitation,to isolate and more importantly,to quantify the impacts of dust aerosol on the vertical structures of clouds and precipitation.To this end,we first investigated the temporal and spatial distribution of Dust Optical Depth(DOD)over global land areas and dust transport patterns over the tropical ocean areas using satellite observations.Moreover,we investigated the influence of dust aerosol on ice cloud top temperature(CTT)by combining multi-sensor,multi-platform satellite observations with the sophisticated cloud resolving model(CRM).We also proposed a new method to quantify the dust aerosol induced changes in CTT of ice clouds.In addition,the potential impacts of mineral dust on rainfall vertical structure were studied by employing the observations of Tropical Rainfall Measuring Mission(TRMM)and the empirical orthogonal function(EOF)analysis method.The main conclusions are as follows:(1)Spatial and temporal variations of DOD.In this section,a global-scale dataset of DOD was derived from the Moderate Resolution Imaging Spectroradiometer(MODIS)Level 2 Deep Blue aerosol products,which have advantage in retrieving aerosol information over high visible-reflecting surfaces.The spatial and temporal variations of DOD over land at global scale were investigated and the active dust sources on Earth were detected by using this global DOD dataset.Also,the DOD over oceans was derived from MODIS Level 3 Dark Traget aerosol products.The dust transport patterns over the tropical oceans were depicted by the EOF analysis.Results show that during the study period(2005-2014),the DOD is inhomogeneously distributed over global land aeres,and shows significant seasonal variations at regional scale.The high values of DOD distribute geographic coherently along the "Dust Belt",which extends from west to east covering the North Africa,the Middle East,the Indian Subcontinent and the East Asia.The North America and Australia are also active dust sources.These above six regions are also characterized by the most frequent and strongest outbreaks of extreme dust events.Furthermore,the seasonal variations of DOD in different dust sources were investigated.The maximum DODs are in boreal summer in the North Africa,the Indian Subcontinent and the North America,while the maximums of seasonal mean DODs in the Middle East and the East Asia are in boreal spring.The seasonal variation of DOD in Australia is insignificant.In boreal winter,the global mean DOD is limited in a relatively low level.Additionally,the inter-annual variations of DOD at regional scale were also investigated.There are decreasing trends of DOD from 2005 to 2014 in the North Africa,the Indian Subcontinent and Australia.However,in the Middle East and the North America,there are significant increases of DOD since 2010,while there is no substantial trend of DOD in the East Asia during the study period.Furthermore,the DOD variabilities over the tropical oceans were decomposed by the EOF analysis.Results show that the EOF analysis is successful in isolating the major dust transportation pathways over ocean in different seasons.(2)Reducing the uncertainties in dust aerosol IN effect on the CTT of ice clouds.Dust aerosol could act as ice nuclei(IN)to mediate the phases and sizes of hydrometers through heterogeneous ice nucleation,leading to significant modification of ice cloud properties.In this study,the potential impacts of dust aerosol on the CTT of ice clouds were investigated by employing multi-sensor and multi-platform satellite observations and cloud resolving model(CRM).Also,a new method was proposed to reduce the uncertainties in satellite observational study of aerosol indirect effect on the CTT of ice clouds.In this study,the horizontal and vertical distribution of cloud and dust aerosol were investigated by using multi-sensor satellite observations including the Moderate Resolution Imaging Spectroradiometer(MODIS)onboard the Aqua satellite,the Cloud-Aerosol Lidar with Orthogonal Polarization(CALIOP)onboard the CALIPSO satellite and the Cloud Profiling Radar(CPR)onboard the CloudSat satellite.The development of this cloud system was simulated by WRF(Weather Research and Forecasting)model without considering any dust aerosol effect.The WRF simulated cloud paremeters(i.e.,CTT and cloud water path)were compared with the satellite observations to ensure that the WRF simulation could capture the main properties of the clouds system.Furthermore,the three-dimensional wind fields inside the cloud systems derived from the simulation were combined with satellite observations to determine the mixing states between clouds and aerosols.These above steps help to minimize the errors in classifying the clouds samples truly mixed with dust aerosols.Futhermore,we examined the sensitivities of 4 different microphysical schemes,3 different initial conditions,3 different cumulus parameterizations and 3 different heretogeneous ice nucleation parameterizations in WRF/WRF-Chem on the simulation of ice cloud CTT.Satellite observations show that the linear regression slope of CTT against logioCIWP was about 5 times steeper than the value in relatively pristine ice clouds.Moreover,the probability distribution function(PDF)of CTT for dust-mixed ice clouds shifts to the warm end and shows an additional peak at-25 ?.For a given cloud ice water path and a given cloud optical depth,the mean CTTs for the dust-mixed ice clouds were systemically warmer than those of the pristine clouds.These facts indicat that there were more ice clouds formed at relatively warmer temperature in the dust-mixed clouds,which definitely attributed to the heterogeneous ice nucleation.Numerical simulation agrees well with the satellite observations in the pristine ice clouds,but fails to depict the CTT variations in dust-mixed ice clouds,in terms of the slope of CTT-log10CIWP and the PDF of CTT.Assuming that the WRF simulation represents the meteorological effects on ice cloud CTT,the pure dust aerosol IN effect on CTT was quantified by subtracting the meteorological effect from the satellite observed total variations of CTT.In this case,the quantified IN effect(i.e.(?)CTT/(?)AOD)ranges from 15? per unit AOD to 35.3? per unit AOD in the dust-mixed clouds.To make the results more robust,WRF simulations with 4 additional microphysical schemes,2 more initial conditions,and 2 more cumulus parameterizations were conducted.Sensitivity tests show that the simulated ice cloud CTT are sensitive to the microphysical schemes and initial conditions,but insensitive to cumulus parameterizations.Generally,all the simulations show large departures with satellite observation in the warm peak of CTT PDF at-25 ?.This indicats that the microphysical parameterizations without taking into account the properties of IN source aerosols(e.g.mineral dust)may yield to large uncertainty in simulations and cannot explain the satellite observations.The sensitivity of the heterogeneous ice nucleation parameterizations shows that Meyers,Demott and Niemand schemes are able to simulate the warm peak of CTT PDF at-25 ?,but shift the peak aroud-43 ?to the cold end.This may indicate that the model fails to consider the influence of initial heterogeneous ice nucleation on subsqequent homogenesou nucleation events.(3)The impacts of mineral dust on rainfall vertical structure as identified by the EOF analysis.In this section,the potential effects of dust aerosol on rainfall vertical structure were investigated by employing the three-dimensional observations of precipitation by TRMM PR and the EOF analysis method.In addition,the long-term correlation between the storm height and the background aerosol cover was studied.Results show that the rainfall vertical structure is determinted by the combined effects from cloud dynamics,thermodynamics,and microphysical processes,and has potential interactions with dust aerosols.For a given PR near surface rain rate,the mean storm height for stratiform precipitation is significantly higher in the dust-laden section than that in the dust-free condition,while such difference is not evident for convective precipitation.The variabilities in vertical precipitation structure were decomposed by the EOF analysis.It was found that the leading three normalized EOF modes(i.e.EOF1,EOF2,EOF3)for convective precipitation and the EOF1 for stratiform precipitation show consistent structures between the dust-laden and the dust-free sections.However,for the EOF2 and EOF3 of stratiform precipitation,there were large departures of rain rate above the freezing level between the dust-laden and the dust-free sections,indicating an enhancement of cold rain processes in the stratiform regimes under dust-laden condition.The physical interpretations of the EOF modes were investigated by the correlation analysis between the EOF Expansion Coefficients with various atmospheric parameters.Results indicate that the EOF2 and EOF3 for stratiform precipitation show enhanced correlations with the column precipitable water and the ice hydrometers amounts,but weak correlations with other parameters.Therefore,it was inferred that the EOF2 and EOF3 for stratiform precipitation might represent the effects of mineral dust in promoting heterogeneous ice production in the stratiform regimes under dust condition.In this case study,the dust-induced changes in rainfall vertical structure account for 6.12%of the total vertical variabilities.The statistical study shows consistent results with the case study,except that the aerosol-related changes in stratiform rainfall vertical structure is isolated in the EOF3 only,accounting for 2.39%of the total vertical variabilities.This study suggests that EOF analysis is a promising way for isolating the potential and relatively weak aerosol effects on precipitation from the strong dynamic effects.Furthermore,a correlation coefficient of 0.512 is observed between the monthly mean normalized SH(NSH)over the Atlantic Ocean(6S-10N and 40W-11E)and the coarse mode AOD in the background.This correlation is comparable to or even stronger than those between dynamical conditions and NSH,indicating that the impact of dust aerosol on the vertical development of storm may have significant climate effects.