| The net aerosol effects on clouds and precipitation are highly debated,largely due to the complex cloud microphysical processes and confounding meteorological co-variations that surround the aerosol-cloud-precipitation interaction.Aerosol effect is mainly considered as twofold:on one hand,aerosol particles can absorb or scatter solar radiation,reduce energy that can reach the surface and lower atmosphere,therefore,adjust the vertical distribution of temperature and water vapor;on the other hand,aerosol particles can act as cloud condensation nuclei and ice nuclei,which can therefore significantly affect cloud systems by changing a myriad micro-physical and related dynamic processes.Both effects of that can ultimately affect the formation and development of cloud systems,and their resultant precipitation properties.As a result,the net aerosol effect on cloud systems is of great importance in both climate modeling and weather forecasting,which is,however,far away from been fully understood.In this dissertation,based on explicit observational data analyses and cloud-resolving model simulation analyses,with foci on deep convective precipitating systems,we investigated possible aerosol effects on various precipitation regimes and their corresponding physical mechanisms.First of all,by using 6 years?2007–2012?of observational data over the Pearl River Delta in southeast of China,including collocated ground-based aerosol information from ground sites,precipitation profiling measurements from precipitation radar?PR?on board Tropical Rainfall Measuring Mission?TRMM?satellite,in combination with re-analysis datasets?ERA-Interim?from European Centre for Medium-Range Weather Forecast?ECMWF?,we developed a novel method to identify localized precipitating events,which was further used to analyze the response of three different regimes of localized precipitation?shallow convection,stratiform and deep convection?to aerosols.As compared to pristine atmospheric conditions,polluted atmospheres were found to relate with weaker rain rate produced by warm precipitating processes,albeit no significant influences exerting on their vertical development.For the mixed stratus precipitation,higher aerosol concentrations tended to inhibit their vertical development,thereby reducing the rain rate.However,higher aerosol concentration fueled the vertical development of mixed-phase convective precipitation,which in turn significantly enhance rain rate.Based on previous studies,we suggest that the above-mentioned aerosol suppression effects on warm precipitation was mainly due to delayed collision-coalescence processes,whereas the aerosol inhibition effects on mixed stratus precipitation was mainly due to the weak updrafts that can hardly bear the additional mass produced by more efficient condensation process under polluted conditions.By comparison,aerosol invigoration effects on mixed convective precipitation was mainly due to additional mass and latent heat releases produced by more efficient condensation process under polluted conditions.Notably,the meteorological effects tended to affect such aerosol effects.For better improvement on understanding of the mechanism by which aerosol affects convective precipitation,the response of deep convection over the tropical oceans to aerosols have been systematically analyzed.In particular,the net aerosol effects on deep convective precipitations were elucidated based on 10 years?2003–2012?of satellite data.Analyses revealed a pervasive boomerang-shaped relationship between aerosol and convective cloud systems.Overall,as aerosol optical depth?AOD?increases,deep convective clouds and their rain rates initially showed sharp increasing trends,which then leveled off or even turn into weakly decreasing trends.This phenomenon was generally termed as“Boomerang effect”.Such boomerang effect was found for different seasons,regions and meteorological conditions.But the optimal AOD(AODop),which corresponds to the maximum R,tended to be larger for stronger precipitating systems.Possible contamination from data retrieval,statistic bias,meteorological co-variations,and wet scavenging were carefully examined.We argued here that such universal and systematic boomerang R-AOD relationships were produced by the competition between more efficient condensation and evaporation processes under higher aerosol concentrations.Further,the suppression on convective clouds caused by aerosol radiative effects,which is proposed as the mechanism for similar boomerang relationships over Amazon,can hardly explain our observations for two reasons:one is the extremely limited aerosol-solar radiation interaction with high cloud fractions that are produced by fully developed deep convective clouds;another is due to the slow temperature responses over ocean surface that are caused by relatively high specific heat capacity of water.For detailed interpretation of such boomerang relationships between aerosol and deep convective systems,we performed explicit simulations,using both the TAU-CM model?for ideal deep convective clouds?and WRF model?for an observed deep convective system?.Different aerosol concentrations and detailed cloud micro-physical processes enabled us to figure out the underlying mechanism of net aerosol effect on deep convective systems.Simulations for ideal deep convective clouds revealed similar boomerang relationships to those revealed in observational analyses,but in terms of cloud mass and surface rainfall.Furthermore,the turning points for these simulated boomerang relationships are of greater aerosol concentrations while considering more humid atmospheric conditions,again.The competition between additional condensation and evaporation due to additional aerosols was exhibited as expected.Similarly,simulations for the observed deep convective system showed much higher condensation rate over cloud-core region,compared to much higher evaporation rate over cloud-periphery region.By and large,deep convective system under polluted conditions tended to have deeper clouds,and stronger rain rate.However,further analyses for life cycles of single clouds developed during the deep convective system,various and complex signals are shown.Generally speaking,under polluted atmospheric condition,only systematic weaker development and stronger negative buoyance at the top of warm convective clouds are clearly found,while mixed convective precipitating clouds show almost no systematic responses.To sum up,using long-term observational data and model simulations,the net aerosol effects on different types of precipitation systems,especially on deep convective systems with large rain rate,were investigated.Our findings show high consistency with previous studies,providing considerable and convincing evidence of solid aerosol influences on precipitating properties.The boomerang effect of aerosol on precipitation systems,albeit with dependence on meteorological conditions,is expected to give unique insight into the mechanism underlying how aerosol affects localized precipitation over highly-polluted region like China and deep convective precipitation over tropical oceans,and thus reduce the uncertainty of aerosol-cloud-precipitation interactions.However,the following shortcoming or caveat for this work calls for further investigations in the future.On one hand,for observational analyses,artifacts induced by data retrieval or statistical bias can still play a role in a certain degree.Confounding meteorological co-variations can not be completely excluded by considering several limited key meteorological variables;The chemical properties and nucleation ability of aerosol particles are desperately needed for better quantify aerosol net effects on cloud systems.On the other hand,for model simulations,idealized or simplified cloud systems and atmospheric conditions,poor understanding of ice-related micro-physical processes,and limited spatial and temporal time resolutions certainly contribute to the discrepancies between simulations and observational data.Therefore,in order to elucidate the net aerosol effects on cloud systems,plenty of efforts from both observations and model simulations are warranted. |
PDF Full Text Request