GUO Xueliang,FU Danhong,LI Xingyu,HU Zhaoxia,LEI Henchi,XIAO Hui,and HONG Yanchao

1Chinese Academy of Meteorological Sciences,Beijing100081

2Institute of Atmospheric Physics,Chinese Academy of Sciences,Beijing100029

Advances in Cloud Physics and Weather Modif cation in China

GUO Xueliang?1,FU Danhong2,LI Xingyu2,HU Zhaoxia2,LEI Henchi2,XIAO Hui2,and HONG Yanchao2

1Chinese Academy of Meteorological Sciences,Beijing100081

2Institute of Atmospheric Physics,Chinese Academy of Sciences,Beijing100029

Thecapabilitiesof cloud-resolving numerical models,observational instruments andcloud seeding haveimproved greatly over recent years in China.The subject of this review focuses on the main progresses made in China in the areas of cloud modeling,f eld observations,aerosol-cloud interactions,the effects of urbanization on cloud and precipitation,and weather modif cation.

advances,cloud physics,weather modif cation,China

1. Introduction

Cloud physics is the study of the physical processes that lead to the formation,growth and precipitationof clouds,and is critical in our understandingof precipitationformation and climate change,as well as improving weather forecasts and the scientif c basis of weather modif cation activities.

The state of cloud physics and weather modif cation has been discussed well from the perspective of laboratory, modeling and f eld experimental studies(e.g.,Cotton,1986; Rogers and DeMott,1991;Orville,1996;Pruppacher and Klett,1997;Bruintjes,1999;Silverman,2001;National Research Council,2003;List,2004;Garstang et al.,2005).The major progresses in cloud physics and weather modif cation in China have also been brief y reviewed(Huang et al.,2003; Fang et al.,2003;Guo et al.,2003;Zheng et al.,2003;Mao andZheng,2006;Yao,2006;Maetal.,2007;GuoandZheng, 2009;Guo et al.,2013).

Cloud-resolving numerical models are of great importance in cloud physics and weather modif cation studies(National Research Council,2003;List,2004;Garstang et al.,2005).Orville(1996)reviewed cloud modeling in weather modif cation,and showed that it was not until the late 1950s that modeling began to take shape(Saunders,1957;Malkus and Witt,1959),and theoretical developments in cloud seeding have come about mainly through the use of cloud models.China started to develop cloud models in the early 1960s (Gu,1962,Xu and Gu,1963),but then unfortunately almost stopped for about ten years from the end of the 1960s to the end of the 1970s.However,intensive and substantial f eld experiments on hail suppression were conducted during this period(Huang and Wang,1980).

The reactivation of cloud modeling studies in the 1980s primarily came about to meet the needs of hail suppression activities.A better understandingof hail formation processes such as hail embryo formation,transport and further growth is essential to conduct effective hail suppression operations. For this purpose,hail-cloud models in 1D,2D and 3D with one-moment and two-moment microphysical schemes have been constructed and applied in studies of hail formation and growth processes in convective clouds(Mao et al.,1982;Xu and Wang,1985,1988;Hu and He,1987),almost in parallel to those studies conducted by researchers outside China in the late 1970s and early 1980s(e.g.,Orville and Kopp, 1977;Weisman and Klemp,1982,1984).Most hail-cloudmodels use elastic dynamics approximations,simplif ed microphysics and ideal initiations.Very few models included cloud seeding processes,as performed by Orville and Chen (1982),until the 1990s,although the effect of cloud condensationnuclei(CCN)oncloudformationhas beeninvestigated in a few studies(e.g.,Xu et al.,1983).

Important progress in the developmentand application of cloud models was made in and after the 1990s,as more powerful computing resources became available.Cloud models with sophisticated dynamics and detailed ice microphysics have been developedto study hail formationprocess in China (Konget al.,1990,Kong,1991,Kongetal.,1991;Guo,1997; Hong,1998).

The AgI-seeding process was developed in a 2D version of a cloud model(Huang and Xu,1994),which was quite similar to the work of Orville and Chen(1982).The nucleation processes of the seeding agents of AgI and liquid-CO2were developed in the hail-bin model to compare dynamic and microphysical effects induced by two seeding agents (Guo et al.,2003;2006a)in an attempt to obtain optimal cloud seeding methods(Guo et al.,2007).However,most cloud models are initialized by a single sounding dataset, which providesa homogeneousmeteorologicalf eld and cannot ref ect realistic cloud formation and evolution.To solve this problem,studies on cloud models initialized with output from the Fifth-Generation Penn State/NCAR(National Center for Atmospheric Research)Mesoscale Model(MM5) were conducted(e.g.,Guo and Fu,2003;Kang et al.,2004a, 2004b).

With thedevelopmentofmesoscalecloud-resolvingmodels[e.g.,MM5,the Weather Research and Forecasting (WRF)model,and The Global and Regional Assimilation and Prediction System(GRAPES)]and powerful computing ability,more sophisticated and detailed dynamics and microphysics are available in these models.The AgI-seeding process was f rst developedin WRF-based cloud-resolvingmodels in China,because the seeding effect on cloud and precipitation can be more realistically tested and evaluated based on this model(Fang et al.,2009).Also,the application of these cloud-resolving models has become common in a variety of topics,such as the formation mechanism of severe storms, aerosol-cloud interactions,and the impact of cloud merging, ice processes,and topography on precipitation(Guo et al., 2006b;Fu et al.,2011;Fu and Guo,2012;Guo et al.,2013), as well as guiding operational cloud-seeding activities(Lou et al.,2012).The studies related to these issues in China are also reported in this review paper.

China has been one of the most active countries in operational weather modif cation programs.The f rst cloudseeding experiments were conducted in both northern and western China in 1958(Cheng,1959).Meteorological disasters account for over 70%of natural damage in China,in which severe drought,f ood,hailstorms and fog are more frequent.The severe situation of weather hazards and water stress has motivated the rapid development of weather modif cation activities in terms of rain enhancement and hail suppression in China.

Important progress in technologies and methodologies of weather modif cation has been made in the past several decades,particularly in the development and application of advanced observational tools in operational weather modif cation activities.An important achievement is that modern technologies such as the Global Positioning System (GPS),remote sensing,and Geographical Information Systems(GISs)havebeeneffectivelycombinedandincorporated into decision-making systems of weather modif cation activities,through which the displaying,monitoring and tracking of cloud seeding are available.The different types of radars, satellites,airborne measurements and operational mesoscale models are widely used in guidingand evaluatingcloud seeding.

This paper reviews the current state of cloud physics and weather modif cation,paying particular attention to the modeling,technological and methodological developments in the last several decades in China.In doing so,the major uncertainties or challenges that limit advances in research and operational weather modif cation are identif ed,and recommendations for future avenues of research are provided to help reduce these critical issues.Following this introduction, section two addresses the progress made in cloud physics and weather modif cation in China,including cloud modeling and simulations,f eld investigations of aerosol,CCN and cloud properties,climatology and variations in cloud water retrieved from satellite data,as well as weather modif cation and its scientif c basis in China.Conclusions and further discussion are presented in the f nal section.

2. Progress in cloud physics and weather modif cation

The complicated formation processes of cloud systems lead to an incomplete understanding of the precise mechanism of how a cloud forms and grows.In addition,different cloud formation mechanisms in different climatic and geographical regions increase the diff culty of cloud formation studies.Over several decades,studies of cloud physics in China have seen increasing interest in a wide variety of topics covering cloud modeling and f eld experiments,relationships betweencloudphysicsand precipitation,hail,lightning,and aerosol-cloud-radiationinteractions,due to the increasing demandsof weather forecasting,climate changeand weather modif cation activities(Ma et al.,2007;Guo and Zheng,2009).

2.1.Clouds and precipitation modeling and simulations

2.1.1.Hail-cloud modeling and simulations

As acommonformofmeteorologicaldisaster,hailclouds cause great damage to crops and property every year in many areas across China.In order to understand hail formation processes and search for effective hail suppression technologies,Chinese scientists havefordecadesmadegreateffortsto understand hail formation mechanisms based on hail-cloud modeling studies(Huang et al.,2000,2003;Zheng et al.,2003;Guo and Fu,2003;Fang et al.,2003).Cloud models have been developed and applied in many aspects.

2.1.1.1. Understanding the microphysics and dynamics of hail clouds

A large amount of published papers can be classif ed into this category(Xu and Xia,1964;Mao et al.,1982;Xu,1985; Xu and Wang,1985;Hu and He,1987;Xu and Wang,1988; Kong et al.,1990,Kong,1991;Kong et al.,1991;Guo,1997; Hong,1998;Lin et al.,2000;Guo et al.,2000,2001a,b; Guo and Huang,2002;Hong et al.,2002;Qi et al.,2002a, b;Xiao et al.,2002;Wang et al.,2002;Fang et al.,2002; Kang et al.,2004a,b).Cloud models have developed from 1D to 2D and 3D with one-moment,two-moment and bin microphysical schemes.Kong et al.(1990,1991)developed a time-dependent,non-hydrostatic,fully-elastic 3D cloud model with a one-moment scheme of hail microphysics and investigatedthe effect of low-level wind shear and cold water surfaces on hailstorm development and movement.The size distribution of hailstones is usually assumed to be distributed by an inverse exponential size distribution and their growth rates are based on mass weighted mean terminal velocities in one-and two-moment schemes.These assumptions may cause large errors,particularly in the size ranges varying signif cantly over the spectrum.Therefore,a cloud model with hail-bin microphysics was developed and applied in hail formation research(Guo,1997;Guo et al.,2000,2001a,b;Guo and Huang,2002).The hail growth via the recirculation process in a multicellular hailstorm can be simulated well.

Hail forecasting and warning systems have also been tested(Mao et al.,1982;Xiao et al.,2002)as previously carried out by Cotton and Boulanger(1975).In addition,the hail-cloud models used in this period were usually initialized with single sounding datasets and the convection was initiated with an idealized thermal bubble,which cannot ref ect the formationand evolutionprocesses of real hailstorms,thus limiting the application in hail suppression activities.In order to predict the location,time and amount of hailfall and providemorerealistic informationonhail clouds,cloudmodels initialized with MM5 have been attempted(Guo and Fu, 2003;Kang et al.,2004a,b).

2.1.1.2.Understanding cloud seeding process

Cloud seeding processes,similar to those in the work of Orville and Chen(1982),have been developed in cloud models and applied in evaluating and understanding AgI-seeding effects(Hong,1999;Li et al.,2003;Zhou et al.,2003;Guo et al.,2006a;Xiao et al.,2006),testing hypotheses such as the role of the accumulation zone of supercooled water in hail formation,as proposed by Sulakvelidze et al.(1967)in the 1960s It has been found that there is an obvious accumulationzoneofsupercooledwater,whichplaysan importantrole in forming hailstones in convective clouds in western China (Zhou et al.,2001;Hu et al.,2003),and the so-called benef cial competition mechanism has been described(Xu and Duan,2001,2002).Also,cloud models have been applied in optimizing seeding methods(e.g.,Guo et al.,2007),leading to the f nding that the seeding in the maximum updraftregion or maximum supercooled water region can obtain the largest surface rainfall.The application of cloud models in weather modif cation is discussed in more detail in section 2.4.

2.1.1.3.Investigating severe wind-producingstorms

Cloud models have also been used to better understand the formation mechanisms of severe winds(downbursts)produced by convective cloud systems.Cloud models with bin microphysics have been used to investigate the contribution of the various hydrometeor types on downdraft formation by calculating the equivalent cooling rates(Hjelmfelt et al., 1989;Guo et al.,1999a,b;Guo and Fu,2003;Fu et al.,2003; Sun et al.,2004;Fu and Guo,2007).

Time-height distributions of the simulated maximum downdrafts and hail/graupel and rainwater content indicate that downbursts are primarily produced by hail/graupel and rain loading in the upper layers,and then enhanced by hail/graupel melting and rain evaporation after the particles pass through the melting layers.

Severe surface winds could not be reasonably simulated and explained by negative buoyancy induced by loading and melting processes in a case simulated by Liu and Guo (2012a).Recent studies found that a descending rear-inf ow jet,which was collocated with a mesovortex along the leading convective line,might cause severe surface winds(Trapp and Weisman,2003;Atkins and Laurent,2009).Houze et al. (1989)proposed the concept of convective momentum transport(CMT)in mesoscale convective systems(MCSs)to illustrate the accelerated f ow,and how f ow from the rear of the MCS can be accelerated via the pressure gradient force and transported downward.

The effects of the surface cold pool and CMT in the formation of severe surface winds on 3-4 June 2009 in China were investigated using the WRF model.The results indicatedthat therewas a strongwesterlyjet belt withwind speed of＞30 m s-1and 5 km thickness at levels of 11-16 km. The jet belt accelerated and descended while the squall line convective system occurred.The negative buoyancy due to the loading,melting and evaporation of cloud hydrometeors induced the downward momentum transport from the upper levels.The contributions to the formation of severe surface winds from downward momentum transport accounted for about 70%,while the surface cold pool accounted for about 30%(Liu and Guo,2012a,b).

Withthedevelopmentofmesoscalecloud-resolvingmodels(e.g.,MM5,WRF,GRAPES),more detailed hail-cloud microphysics has become available.The application of these mesoscale models in cloud and precipitation formation studies,as well as inguidingcloud-seedingactivities,has become increasingly common in China.The relevant progress in this regard is described in the following sections.

2.1.2.Effects of ice microphysics and cloud-merging processes on precipitation formation in MCSs

Various effects of cloud microphysics on the formation and evolution,as well as the precipitation,of MCSs havebeen described.The conclusions,however,are inconsistent, and even contrary to one another in some respects.When convectivesystems are simulated without ice microphysics,a reduced stratiform component can occur,as well as a faster propagation speed and a shorter life cycle,as compared to simulations with ice microphysics(e.g.,Grabowski,2003; Tao,2007;Tao and Moncrieff,2009).On the contrary,similar propagation speeds and life cycles of convective systems have also been obtained with and without ice microphysical simulations(e.g.,Tao and Moncrieff,2009).

Investigationsontheprecipitationdifferencebetweenicefree and ice runs have also been conducted,and based on the Tropical Ocean Global Atmosphere Coupled Ocean Atmosphere Response Experiment(TOGA COARE)bulk f ux algorithm it has been found that precipitation increases in the ice run(e.g.,Tao et al.,2003).The recently improved ice microphysical processes in the WRF model enable a more realistic representation of the formation of cloud ice in the cloud model(Hong and Lim,2006).The effects of cloud microphysics on monsoon MCSs was investigated by Fu et al. (2011)based again on the cloud-resolving WRF model over the South China Sea Monsoon Experiment(SCSMEX)region.The sensitivity experiments with ice runs and ice-free runs showed that the ice run can produce less rain prior to and during the onset of the monsoon period.The reason is due to the different cloud microphysics having different responses to radiation processes.Therefore,the effect of ice microphysicson precipitationis determinedthroughthe radiation processes instead of the microphysics itself.

The role of cumulus merging processes in the formation of MCSs has been widely investigated in terms of the development and enhancement of the cloud echo area,mixedphase microphysics,rainfall and cloud-to-groundlightingactivities,as well as the possible application in weather modifcation(e.g.,Rosenfeld and Woodley,1993;Changnon et al., 1995;Czys et al.,1995).Previous numerical simulations, however,generally adopted idealized methods to initiate the convection,which could not objectively ref ect the natural cloud-merging processes.Fu and Guo(2006,2012)investigated the cumulus-merging processes in generating an MCS that occurred on 23 August 2001 in the Beijing region by using a cloud-resolving mesoscale model of MM5,and found that the formation of the MCS experienced multiscale merging processes from the single-cell scale to the cloud clusterscale,and echo-cores merging within the MCS.The merger process can apparentlyalter cloud dynamicaland microphysical properties through enhancingboth low-and middle-level forcing.Also,lightning f ash rates are enhanced by the production of more intense and deeper convective cells by the merger process,by which the more graupel-like ice particles especially are formed in clouds.Explosive convective development and a late peaking lightning f ash rate can be found during the merger process.

Li et al.(2011,2013)investigated the cloud structure and evolution of MCSs retrieved from the Microwave Imager(TMI)and Precipitation Radar(PR)of the Tropical Rainfall MeasuringMission(TRMM)andmadecomparisons with some pioneeringstudies based on soundingsand models over the northern South China Sea.TRMM-retrieved rainfall amounts were generally consistent with those estimated from soundings and models.However,sounding-and PR-based estimates were relatively higher than TRMM-TMI-retrieved results.The WRF-based simulation underestimated the maximum rainfall rate by 27%compared to that derived from TRMM-PR,and underestimated mean rainfall by 10.4%and 12.5%comparedto the TRMM-TMIand soundingestimates, respectively.In addition,apparent phase shifts among different retrievals existed.

2.1.3.The effects of modif cation of urbanization and aerosols on cloud and precipitation

Many observational and modeling studies outside China have indicated that rainfall patterns,amount and distribution in and downwind of cities can be modif ed(e.g.,Bornstein and Lin,2000;Shepherd et al.,2002;Rozoff et al.,2003). In China,abnormal precipitation distributions and intensities have often been found in large cities such as Beijing and Shanghai,which poses a new challenge in forecasting severe convective systems in urban regions due to the rapid development of urbanization.Guo et al.(2003,2006b)studied the effect of urbanization on MCSs in the Beijing region using the MM5 model,and found that total precipitation in the urbanized region decreases,and its distribution tends to become concentrated and also intensif ed along the borderline between the urban and non-urban region.The precipitation intensity is also modif edand locally distributed,and the high precipitation intensity core is found downwind of the city. This study is more consistent with those observed in this region.Thus,the urban region could act to create a bifurcation zone for precipitation distribution and produce more f oods, as suggested by early studies.

The majority of the aforementioned studies focused on the dynamic forcing mechanisms related to urban land-use or change and did not address the potential role of aerosols. The aerosol particles in the cloud-forming air are one of the key factors in determining cloud and precipitation formation, which act as CCN or ice nuclei(IN).Thus,studies about the relations between aerosol and CCN are very important. With the increase of industrial air pollution and rapid urbanization in China,studies on aerosol-cloud interactions have also been conducted in recent years(e.g.,Huang et al.,2003; Lou et al.,2003;Yao,2006;Ma et al.,2007;Duan and Mao, 2008;Lei et al.,2008;Guo and Zheng,2009).Furthermore, the effects of CCN on cloud formation in China have been modeled(e.g.,Xu et al.,1983;Xiao et al.,1988;Zhao et al., 1998),and the impacts of sea salt and non-sea-salt sulfate on marine cloud microphysical properties investigated(Zhao et al.,2005a,b).

A cloud-resolvingmodel includingaerosol processeswas used to study the effects of aerosols on the dynamics and microphysics of cloud(Fang,2008),in which it was found that the increase of aerosol concentration suppressed the updrafts and downdrafts,and resulted in more(but smaller)cloud drops and more cloud water,which in turn lead to an in-crease in albedo.Smaller cloud drops have less eff cient at coagulating to form raindrops and thus lead to less raindrop growth.For the whole cloud system,therefore,the impact of increasing aerosols inhibits precipitation.Guo et al.(2014) modeled aerosol impacts on summer convective clouds and precipitation over northern China by using the WRF model coupled with Chemistry(WRF-Chem)and found that the domain-averagedprecipitation amount under polluted conditions can be increased up to 17%during the whole cloud lifetime.However,the maximum rainfall rate above 30 mm h-1is enhanced,whereas that below 30 mm h-1is suppressed in most cloud lifetime.The differences of cloud microphysics and dynamics between pollutedand clean conditions indicate that both warm and ice microphysics and updraft are suppressed at the storm’s initial and dissipating stages,whereas those at the storm’s mature stage are obviously enhanced under polluted conditions.

Water-soluble dust particles can increase the concentrations of giant CCN,leading to earlier development of icephase precipitation particles,thus promoting the formation of precipitation(Yin et al.,2000;Chen et al.,2007;Yin and Chen,2007;Chen and Yin,2008).In addition,IN are very important in many weather events for the reason that IN can affect the initial concentrations of ice particles in cold clouds and then change the physical characteristics of cold clouds, which may inhibit the developmentof ice-phase precipitation particles(Yin et al.,2002;Chen and Yin,2008,2009).

WRF-based modeling studies have found that urbaninduced aerosol pollution can result in a decrease in rainfall over the urban area,with the suppression occurring particularly in the downwind area(Fig.1).

In addition,sensitivity studies using the WRF-Chem model have been conducted to investigate the effects of secondary aerosols formed by gaseous emissions of SO2and NO2on precipitation,and the results indicate that secondary aerosols can affect the intensity,duration and location of precipitation.Sulfate aerosol can increase the number of cloud droplets and enhance convection,and accelerate the formation of rain in the high cloud-water content region.In the low cloud-water content area,both SO2and NO2can inhibit precipitation,and NO2is the more eff cient.Increasing SO2concentrations could lead to a dramatic effect on precipitation,in which the strong precipitation process is enhanced while the weak precipitation process is suppressed.

Recently,Jia andGuo(2012a,b)investigatedthe effectof pollutants on foggyevents in China based on the WRF-Chem model coupled with local anthropogenic atmospheric pollution emissions,and indicated that the simulated fog distribution and intensity with anthropogenic emissions were closer to those observed.Polluted atmospheric boundaryconditions were found to create favorable conditions for fog formation, increasing the fog area up to 50%,the maximum intensity by fve times,and the duration to 1.5 h on average.The sensitivity studies showed that the secondary aerosols of SO4, NO3and NH4formed from pollutant gaseous precursors of SO2,NOxand NH3had important impacts on the formation processes and microphysical structures of the foggy events.

2.2.Field investigations of aerosols,CCN and cloud properties

Field investigation of aerosols,CCN and cloud properties are important to reveal the connections of the background aerosol concentration,size distribution and chemical composition to clouds and precipitation formation processes, which is also critical for understanding cloud-seeding processes.China has conducted intensive f eld investigations on aerosols,CCN and cloud properties in the past several decades by using cloud-seeding aircraft equipped primarily with airborne particle measuring system(PMSs)or DMT. Airborne cloud measurements used in weather modif cation operations have developed rapidly in China.These probes have become the principal tools for characterizing aerosol and cloud particle properties,providing more detailed information on clouds and helping to improve weather modif cation operations.

Before 1980,Mee-130 instruments were used to measure CCN properties(e.g.Fan and An,2000;You et al.,2002a). After 1980,the PMS was f rst introduced to China due to the need for cloud-seeding activities,but it was also used for observations of aerosols and cloud structures in both northernand western China.Observational results have been reported in several papers(e.g.,Li et al.,2003;Liu et al.,2003,2005; Su et al.,2003a,2003b;Wang and Lei,2003;Wang et al., 2005;Yang et al.,2005;Jin et al.,2006;Li et al.,2006),and aircraft-based investigations of aerosol size distribution under cloudy and cloud-free conditions have been conducted in many regions of China(e.g.,Zhou et al.,2004;Zhang et al., 2006;Zhao et al.,2006a;Fan et al.,2007,2010;Qi et al., 2007;Zhang et al.,2007,2011).

The observed data over northern China indicate that the CCN concentration is fve times higher over urban areas than over rural villages(e.g.,Shi and Duan,2007;Duan and Mao, 2009).Observations in Beijing have shown that the concentrations of IN,which could be activated at-20°C,increased by about 15 times from 1963 to 1995(You et al.,2002b).

The vertical distribution of aerosol number concentration and diameter sampled with the Passive Cavity Aerosol Probe (PCASP)on two cloud-freedays in Beijing City showedthat, under clear conditions,there was a strong accumulation zone of haze aerosols near the surface on both days.However, the number concentration on 12 September 2004 was much higher than on 29 August 2004 due to the effect of the strong inversionlayer.The higher values of aerosol number concentration and diameter at 6000-6500 m were primarily due to the effect of upper-level dust f ow in the northwestern region.

Lu and Guo(2012)investigated the aerosol distribution, sourceandrelationshipwith CCN basedontheBeijing Cloud Experiment(BCE)and indicated that,due to the different sources of aerosol,the conversion ratios of aerosol to CCN were less than 20%below 4500 m,and reached 50%above the level at 0.3%supersaturation(Fig.4).The aerosols at higherlevels(above4500m)were stronglyaffectedby largesizedparticlesandthosebelow4500mwerestronglyaffected by local or regional pollution.

Measurements of microphysical properties of stratus cloud carried out by an airborne PMS in Hebei Province in October 1990 and April 1991 found that there is a positive correlationbetweenaerosolnumberconcentrationsbelowthe cloud base and cloud droplet number concentrations(Huang et al.,2005).

Aspects of cloud microphysics observed with PMS probes on four aircraft f ights during August and September in 2003 in Beijing and surrounding regions showed that the maximum number concentration of small particles sampled with the FSSP-100 probe changed from 120 cm-3for stratocumulus clouds to 183 cm-3for deep altostratus clouds, and the mean diameter was around 7.22-16.05μm(Zhang et al.,2007).The data of the 2D-GA2 probe showed that the maximum number concentration of large particles changed from 2.25×10-3cm-3to 3.29×10-1cm-3.The maximum water content explored with the King-LWC probe was 0.42-0.69 g m-3in stratocumulus clouds.

Both the vertical and horizontal distribution in stratiform clouds showed clear inhomogeneous features;high number concentrations of small particles of＞120 cm-1existed in the upper levels(above the-10°C level),but also with some larger sized particles(maximum size reaching 20μm).The liquid water content decreased slightly with height and also kept to around 0.1-0.2g m-3.The layers between-5.9°C to -8°C were composed of a large number of column ice crystals and a small number of riming particles.Those between -8°Cto-12°Cwerealmostrimedparticles,andthoseabove -20°C were primarily dentritic ice particles.The number concentration of large-sized ice particles was around 0.01-1 L-1.The layer-averaged size distributions of cloud particles were quite different in stratocumulus and stratus cloud.

The large particles followed a unimodal distribution in stratocumulus cloud,and the width of the spectrum reached approximately 1500μm.A low number concentration corresponded to a low layer of cloud.However,layer-averaged size distributions in stratus cloud was quite different;for particles less than 400μm in diameter,the size distribution averaged over the layers of 0°C to-8°C and-8°C to-12°C were almost the same,while for particles larger than 400μm in diameter,the size distribution averaged over the layers of -8°C to-12°C had an obvious bimodal-type distribution, and that over the layers of 0°C to-8°C had a multi-modaltype distribution.The width of the spectrum was around 1300-1400μm.

Anthropogenic aerosols might contribute signif cantly to the observed reduction of precipitation over northern China, and provide a possible feedback cycle of aerosol loading and precipitation that produces considerable harmful impacts on air quality,the hydrological cycle,crops,and other environmental problems.Statistical analyses of historical precipitation and aerosols data have revealed that deeper precipitation clouds can be inf uenced by aerosols in the form of precipitation suppression.In particular,the suppression effect is stronger over mountainous areas than over plains,and the inf uence of anthropogenic aerosols on convective precipitationpossiblyplaysanimportantrolein summerovernorthern China(Zhao et al.,2006b;Dai et al.,2008;Duan and Mao, 2009).

In addition,satellite[e.g.Moderate Resolution Imaging Spectroradiometer(MODIS)]data have also been used to study aerosol-cloudinteractions,and have been successful in obtaining aerosol optical depth and cloud effective radius information(e.g.Chen et al.,2004;Zhao et al.,2006b).

Li and Mao(2006)analyzed cold cloud ref ectivity in China during the period 1982-1999 using the PAL (NOAA/NASA Pathf nder AVHRR Tiled Land Data),and foundthat thecold cloudref ectivitycouldhavechangedover some areas during this period,and,in the case of the Beijing area,cold cloud ref ectivity could have changed nonlinearly under the inf uence of IN.

By analyzingsatellite datafromtheInternationalSatellite CloudClimatologyProject(ISCCP),MODIS,andtheClouds and the Earth’s Radiant Energy System(CERES),Huang et al.(2006a,2006b)found that,on average,the ice cloud effective particle diameter,optical depth,and ice water path of cirrus clouds under dust-pollutedconditionsare 11%,32.8%, and42%less,respectively,thanthosederivedfromiceclouds in dust-free atmosphericenvironments;and a signif cant negative correlation exists between dust storm index and ISCCPcloud water path,and the semi-direct effect may play a role in cloud development over arid and semi-arid areas of East Asia and contribute to the reduction of precipitation.

2.3.Climatology and variation of cloud water retrieved from satellite data

Cloud water is a key component in cloud physics and weather modif cation.Although the critical role of cloud water has been widely recognized,its quantitative analysis has beenlimitedbecauseofinsuff cientobservations.Surfaceobservations of cloud have served as the primary source of information about clouds in the past.However,these surface datasets do not really provide enough information on cloud properties,such as cloud water and cloud optical thickness. Owing to the expense and diff culty of observations,f eld experiments only provide short-term cloud water data,but have nevertheless extended our knowledge about cloud processes and helped to validate numerical models.

Measurementsofcloudpropertiesfromsatellites cansupplement surface observations andin situstudies by extending their results to larger spatial and temporal scales,as well as by providing more cloud parameters,including cloud water. The ISCCP was established in 1982 as the f rst project of the World Climate Research Program(WCRP)to collect and analyze a globally uniform satellite radiance dataset to produce new cloud climatology.With the implementation of ISCCP, climatic research on cloud water,presented as cloud water path(CWP)in the ISCCP,as well as the hydrological cycle involving CWP,became possible.

Yu et al.(2001,2004)analyzed the cloud amount and cloud optical depth over southeastern China using ISCCP D2 data and found that the low-and mid-cloud maxima of both cloud amount and cloud optical depth occur in winter and early spring,and suggested the effect of the Tibetan Plateau on the westerly f ow as being the cause of different cloud characteristics.In winter,westerly f ow prevails and covers the Tibetan Plateau.The plateau forces the westerly f ow to go around it in the low levels,and then the two parts of the circumnavigating plateau f ow converge downstream.In the middlelevel,thesurfacefrictionoftheTibetanPlateaucauses divergence(Yu et al.,2004;Li et al.,2004,2005).As a result,large-scale lifting occurs on the lee side of the plateau and results in the formation of stratiform cloud in southeastern China.

Li et al.(2008)studied the climatology and variation of cloud water using ISCCP D2 data.The spatial distributions of annual mean CWP derived from ISCCP D2 and precipitationderivedfromGPCP V.2fortheperiod1984-2004exhibit broadsimilarity in China(Fig.2).Theclose relationbetween cloud water and precipitation is in many ways tied to largescale dynamical patterns and topographical features,and the differencein the SichuanBasin exhibitsa distinct mechanism in cloud formation and the cloud water cycle.

The seasonal means of cloud water ref ect the change of circulation over China.In winter,the CWP distribution pattern is similar to that of the annual mean,with great contrast between northern China(northeastern and northwestern China)and southern China(southeastern and southwestern China).The area of high CWP values covers almost all of southeastern and southwestern China.In northeastern and northwestern China,CWP is less than 0.03 kg m-2due to cold temperatures and insuff cient water vapor transport.In summer,the water vapor brought by the summer monsoon circulation largely affects southwestern China,and the average CWP of this region rises in general.The minimum CWP is still held in northwestern China because of the blocking effect of the Tibetan Plateau.

The transition of circulation from westerly f ow to monsoon circulation can be distinguished from the low CWP values in southeastern China,because in summer the westerly f owshiftsnorthwardandthereisnolarge-scalelifting,which favors the formation of clouds.In spring and autumn,the CWP distributions are under the inf uence of the respective previous seasons.These two seasons share a similar distribution pattern in northeastern and northwestern China.In southeastern China,an apparent difference in CWP distribution between spring and summer is found,through which the weakening topographical inf uence and the enhancing monsoon circulation inf uence are identif ed.In autumn,the sudden shift of the westerly and easterly f ow results in the in-crease of CWP on the lee side of the Tibetan Plateau and the decline in northeastern and northwestern China.

By calculatingthe linear trendof annual meanCWP from 1984-2004(Fig.3a),an increase in CWP can be found in China,with the largest increases occurring in the eastern Tibetan Plateau and the western part of northeastern China, where the linear trends are about 0.015-0.025 kg m-2(10 yr)-1.Over the south part of Xinjiang,mainly in the desert region,CWP decreases by about 0.005 kg m-2(10 yr)-1. The rates of increase over southeastern China and northwestern China are relatively small.Considering the differences in CWP distribution,percentage trends have also been examined(Fig.3b).The largest percentage linear trends,larger than 32%(10 yr)-1,are found in the Tibetan Plateau and the western part of northeastern China.Decreasing trends in CWP are found in the desert areas in the western part of northwestern China and adjacent areas.

From the above analyses of the linear trend,the primary regions with increasing trends are located in the eastern Tibetan Plateau and adjacent areas.These increasing trends of CWP are related to the water vapor variation in the Bay of Bengal,which is the main water vaporsource in China.Trenberth et al.(2005)analyzed the water vapor in four datasets (SSM/I,ERA-40,NCEP-1 and NVAP)from 1988-2001 and found an increasing trend in the Bay of Bengal in all of them. Zhai and Eskridge(1997)calculated the linear trend and percentage linear trend of water vapor from 1970-90 using Chinese radiosonde data,and obvious increases were found in the eastern Tibetan Plateau and southwestern China.According to these facts,the increases of CWP can be reasonably linked to the increases of water vapor.

In summer,the largest increase in CWP in China occurs in the eastern Tibetan Plateau.In northeastern China and the northern part of Xinjiang,CWP decreases by about 5% and 10%(10 yr)-1,respectively.Wang(2001)and L¨u et al. (2004)showedthatthesummermonsooncirculationhasbeen weakening since the end of the 1970s.During the weakening periodofthesummermonsooncirculation,thelocationofthe subtropical high has been farther south and induced circulation that barely transports any water vapor westward through the mountains in the northern part of Xinjiang.Insuff cient water vapor and the blocking effect of the mountains are the possible reasons for the decreasing trends in northwestern China.As for northeastern China,decreasing water vapor transport caused by the same factors can explain the negative CWP trend in summer.The weakening summer monsoon circulation has also caused the drought over North China and f ooding along the middle and lower reaches of the Yangtze River after the 1970s.A relatively large trend of increase occurs in winter.The largest linear trend is located mainly in northeastern China,characterized by a 60%(10 yr)-1rise. In southeastern China,CWP increases slightly by less than 15%(10 yr)-1.The increasing trends of CWP in winter are consistent with the trends of water vapor in the distribution pattern(Zhai and Eskridge,1997),which is considered to be caused by the rises of temperature in winter.

2.4.Cloud-seeding activities and its scientif c basis in China

2.4.1.Current status of operational weather-modif cation programs

Operational weather-modif cation programs,which primarily involve cloud-seeding activities aimed at enhancing precipitation or mitigating hail fall,exist in 30 provinces across China.The fundingthat supports these operationalactivities is primarily from local government,but central government and some enterprises such as hydroelectric power, tobaccoandothereconomiccropcompanies,havealso begun to support weather-modif cation operations in recent years. China now mainly conducts glaciogenic cloud seeding rather than hygroscopic cloud seeding.In recent years,China has alsobegunexperimentingwithhygroscopiccloudseedingusing hygroscopic f ares,but this approach is not yet employed operationally.

A cloud-seeding activity can only be successful if the right amount of seeding material is delivered to the right location at the right time,and for the right time interval. Therefore,the development of highly reliable seeding tools is essential.A relatively rapid advancement in cloud-seedingtools,including aircraft,rockets,artillery and ground-based generators,has taken place in the last decade in China(Fig. 4).

Well-equipped advanced aircraft cloud seeding is mainly organized and conducted by provincial weather modif cation off ces.Stationary and mobile AgI-rocket launching systems have been widely used in weather modif cationactivities,and artillery-seedingwith AgI-shells is still used for hail suppressionin some regions.The rocketsusedforcloudseedingusually contain AgI of 10-15 g,while ground-based generators are used in some mountainous regions in China.Although ground-basedgeneratorshave many advantages,such as high safety levels and a low inf uence on air traff c,uncertaintyexists with this approach insofar as whether or not the released seeding agent can actually reach the intended target.

2.4.2.Operational cloud-seeding guidance and evaluation

The scientif c guidance and evaluation of cloud seeding are key components in operational weather-modif cation activities.Most provinces in China have established weather modif cation operational systems,which include basic information from meteorological networks such as radar,satellite,sounding,automatic meteorological and rain gauge stations,as well as position information from seeding tools to guide cloud-seeding operation.In recent years,GPS has beenwidelyusedinaircraftcloud-seedingactivitiesinChina. The combination of GPS with remote sensing data(e.g.meteorological radar,satellites etc.)and GIS greatly improve the cloud-seeding ability and decease the blindness of cloudseeding activities.GPS,remote sensing and GIS techniques for tracking and guiding cloud seeding have become increasingly popular in cloud-seeding activities in China.Accordingly,weather-modif cation experts can guide aircraft cloud seeding from ground-based real-time cloud information observedbyradarandsatellitesetc.,insteadofdoingsoonboard the aircraft.

One of the most advanced and popular instruments to guidecloudseedingis the Dopplerradarnetwork,whichcovers the major regions of the country.For example,the radar mosaic system used in Beijing weather modif cation was developed based on seven radars in and surrounding Beijing City(Fig.5).The mobile polarization Doppler radar is also starting to be used in cloud observation and weather modifcation experiments(Fig.6).These capabilities are of great value in identifying the types of particles in a cloud,and areofpotentialapplicationinassessingcloud-seedingexperiments sincethe particleclassif cations derivedfromthis radar can be used in many aspects of weather modif cation,such as to reveal the transformation of supercooled liquid water droplets to ice crystals in glaciogenic seeding and the developmentof large drops in hygroscopicseeding,and to identify hailstorms and non-hail storms in hail-suppression seeding.

The current polarization radars,including C-band and X-band,are available in national cloud and weather modif cationresearchprojectsandalso in someprovincessuchas Beijing,Anhui and Hubei,but will expand rapidly in numbers inoperationin the future.The millimeter-wavecloudradarwith wavelengths of 8 mm has also begun to be used in cloud observation,but not popularly at present.

Retrieval information suitable for cloud-seeding operationsfromdifferentsatellites andground-basedGPS-network receivers has also progressed rapidly in recent years(Ma et al.,2007).Satellite retrievals have successfully been used to identify cloud-seeding effects in China(Rosenfeld et al., 2005;Yu et al.,2005).NOAA Advanced Very High Resolution Radiometer(AVHRR)images have revealed conspicuous tracks of glaciated cloud in thick supercooled layer clouds induced by cloud seeding over central China,which is the third and most detailed report of the effects of advertent cloud seeding for precipitation enhancement being detected and analyzed based on satellite observations,and it opens new possibilities of using satellites for directing and monitoringweather modif cation experiments and operations (Rosenfeld et al.,2005).

In some provinces,cloud-seeding aircraft have advanced airborne PMS instruments,which can directly explore the backgroundaerosol size distribution and cloud structure,and determine suitable seeding conditions.More than 10 cloudseeding aircraft have mounted PMS(DMT)instruments in China.

An additional instrument equipped in Chinese weathermodif cation operations is the ground-basedand airbornemicrowave radiometer(e.g.,Lei et al.,2003;Jin et al.,2004). Sincethedual-channelmicrowaveradiometercanretrievethe path-integrated total amount of liquid water and water vapor along its beam by simultaneously measuring emissions from vapor and liquid at frequencies near 21 GHz or 23 GHz and 31 GHZ,it has a more importantapplication in weather modif cation.However,this instrument is not eff ciently used in locating regions of high supercooled liquid water in natural clouds in weather-modif cation operations due to a lack of validation and improvementsof relevant retrieval techniques. More sophisticated ground-based multi-channel microwave radiometers with capabilities in observing water vapor,liquid water,temperatureand humidityhave been used in recent weather-modif cation operations(Fig.7).

Cloud-resolving models play an important role in cloudseedingactivities.Thesemodelscanbeusedinmanyaspects, such as seedingdesign,seedingguidanceand seedingevaluation.At the NationalWeather Modif cationCentre(WMC)of the China Meteorological Administration(CMA),real-time operationalweather modif cationmodels based on mesoscale cloud-resolving weather forecasting models such as MM5, WRF and GRAPES have been established to guide weathermodif cation operations(Lou et al.,2012).These models are modif ed to predict and output information on cloud structure and evolution,such as supercooledcloud water etc.(Fig. 8),or are coupled with AgI nucleation processes,which is import to judge cloud-seeding potential and evaluate seeding effectiveness.Some provincesare also operatingone of these models to guide local cloud-seeding operations.

2.4.3.Scientif c basis for glaciogenic seeding

Glaciogenic seeding is the process of enhancing ice content in clouds either by nucleating new ice crystals or freezing supercooled cloud droplets.The current seeding agents for glaciogenicseeding in China are silver iodide(AgI),solid or liquid carbon dioxide(dry ice),and liquid nitrogen.

The seeding material is one of the most important factors to determine effective cloud seeding.Silver iodide and dry ice have been the most widely used in f eld projects and the most thoroughly tested in the laboratory,yet some aspects of their ice nucleation behavior as well as the interaction mechanisms between the microphysics and dynamics of a seeded cloud are still not completely known.The important advantage of the coolant agent is that the number of generated ice crystals is nearly independent of the temperature.By dry-ice seeding,a vertical generated ice crystal plume tends to cause an organized rapid upward motion of the plume(e.g.,Kraus and Squires,1947)due to vertical integration of the buoyant force.However,dry-ice seeding has a dynamic shortcoming due to its rapid fall speed,which reduces the ice crystal growthtimebyfastupdraftdevelopmentwithinverticallyoriented ice thermals(Schaefer,1946).The number of ice crystals formed by the nucleation of AgI increases drastically as the temperature in the atmosphere decreases(e.g.,Vonnegut, 1947;Garvey,1975),and also suffers fromweak dynamics at low altitude due to a lack of ice nucleation.Since dry-icepellets have a rapid fall speed and have to be dropped from high altitudes,Fukuta(1996a,b)suggested a method to seed the liquid-CO2horizontally at the lower level of the supercooled portionofclouds.His originalphysicalideaoftheliquidCO2is to maximize the interaction time of the rising thermal produced by seeding with the abundant supercooled water at the lower level of clouds and to induce a larger dynamic effect.

Many Chinese researchers have investigated the AgI-seeding effect on cloud dynamics and microphysics by using cloudmodels(e.g.,Youet al.,2002b;Fang,2004;Chenet al., 2005;Fang et al.,2005a,b;Li et al.,2006;Hong and Zhou, 2006;Guoetal.,2006b,2007;Huetal.,2007).Thesemodels have parameterized microphysics incorporated with the AgI nucleation process.The results have shown that cloud seeding may lead to changes in the dynamicsand microphysicsof a cloud.

We also compared the effect of the liquid-CO2seeding technique proposed by Fukuta(1998)with AgI-seeding on cloud dynamics,cloud microphysics and precipitation evolution by developing a 3D cloud model with nucleation processes of liquid-CO2and AgI.The results showed that the seeding by liquid CO2and AgI at the cloud levels of-15°C to-20°C has almost the same dynamic effect on the simulated clouds.The seeding is able to induce the formation of weak convective cells in both seeded and unseeded regions due to latent heat released by the transformation from liquid saturation to ice saturation.However,the initial seeding conducted by liquid CO2in the region of maximum supercooled water with a temperature of 0°C to-5°C enables the production of a much stronger dynamic effect and precipitation by forming many new convective cells at low levels in the later stages of seeded clouds.Twin-rotatingrings are initially formedat lower altitudes of about 4-6km abovegroundlevel in the downstream region near the seeded area.Owing to the effect of stable upper levels,the rings move horizontally along the levels of 4-6 km.By 60 min,the most prominent feature is that the up-and downdrafts meet at around 4-6 levels and force the air to move horizontally,and another obvious ring is formed in the upstream region.This process is favorable for the transition of water vaporto cloud vapor,and also the glaciation of suppercooled water into ice.

Using the same cloud model with the AgI seeding process,sensitivity experiments were conducted to look at responses of cloud dynamics,microphysics and precipitation to different seeding methods with different seeding locations,times and agent amounts.The results showed that the cloud microphysical variations responding to different seeding methods differ greatly.The seeding leads to a decrease in the formation of cloud ice,and it enhances snow formation in the cloud development and maturation stage.It is recommended that cloud seeding conducted in the regions of maximum updraft and supercooled water in convective cloud may maximize the augmentation of surface precipitation.

To better understand the variation of cloud and precipitation induced by seeding in a mesoscale manner,the cloud seeding processes in a cloud model(Guo et al.,2006b)were incorporated into the WRF model to study the mesoscale cloud and precipitationdevelopment,and focus on the targeting of seeding agents,the diffusion and transport of seeding material,the dynamic and static effects on the cloud system, and evaluating precipitation on the ground induced by seeding.The results showed that the supercooled cloud water is almost depleted in the seeded region and increased above it due to dynamical effects.The ice number concentration increases rapidly.The seeding leads to a decrease of graupel and enhances snow formation at the same time.The seeding also enables a redistribution of the accumulated precipitation at the surface and a change in the total precipitation(Fang and Guo,2007;Fang et al.,2009).

2.4.4.Scientif c basis for hygroscopic seeding

The aim of hygroscopic seeding is to promote water droplet growth through coalescence and thereby improve the eff ciency of the rainfall formation process.Earlier hygroscopic cloud seeding in China was based on the direct use of salt powder.However,due to the need for huge amounts of salt powder for each cloud seeding operation,and also the harmful effects on the aircraft itself,hygroscopic seeding ceased in China until recent years.

Statistical,observational and modeling results for larger particles(e.g.,＞10μm diameter)have provided some evidences for hygroscopic seeding(Silverman and Sukarnjanasat,2000).Field experiments or operations,such as those in South Africa and Mexico in which hygroscopic seeding f ares were employed,have shown remarkably similar statistical results in terms of the differences in radarestimated rainfall for seeded versus non-seeded groups(Silverman and Sukarnjanasat,2000),and an increase in rain mass 30-60 minutes after seeding was found.A delayed response in radar-derived storm properties is a possible functionofseeding-induceddynamicprocessesbeyondtheclassical cloud physics results that links cloud condensation nuclei and droplet spectra to rain production(WMO,2000).Hygroscopic seeding experiments have provided statistical support for rainfall increases due to seeding based on single cloud analyses,butthephysicalprocessesleadingtotheseincreases in precipitation are not well understood.

The size of seeding particles is considered to play a critical role in hygroscopic cloud seeding and warm rain enhancement since the cloud condensation nucleation and coalescence processes are found to closely relate to the size of aerosol particles.Johnson(1982)reported that giant(＞1 μm in radius)and ultragiant(＞10μm in radius)aerosol particles will produce a tail of large drops in the cloud-droplet distribution,and then these drops will signif cantly accelerate the development of precipitation via the coalescence process.Yin et al.(2000)showed that the coalescence between water drops can be enhanced by adding giant cloud condensation nuclei,and foundthat the most effectiveparticles werethose with radii larger than 1μm,especially those larger than 10μm.Zhang et al.(2006)suggested that the role of giant CCN on precipitationformationin warm,stratiformclouds is only important when the concentrations of CCN larger than 5μm are high.However,some studies have shown that particle size might not be essential to the formation of raindrops for some clouds,such as a maritime cloud(Takahashi,1976; Takahashi and Lee,1978).

Owing to the inconsistent conclusions obtained in previous studies,further investigations are necessary by designing more detailed numerical experiments,as well as f eld measurements(Orville,1996;Bruintjes,1999;Ma et al.,2007; Guo and Zheng,2009).

The effect of hygroscopic seeding was studied based on 1D cloud model,and it was shown that seeding can apparently alter cloud spectra and enhance precipitating particle formation.The effect of different sizes of seeded particles on precipitation is shown in Fig.9.It shows that different sizes of seeded particles have different impacts on precipitation formation,in which the seeding of larger-sized particles (e.g.,particle sizes larger than 2μm in diameter)has a more obvious effect on precipitation formation.

To further clarify the effect of hygroscopic seeding,a WRF-based mesoscale cloud seeding model was developed to investigate the effects of giant particle cloud seeding on cloud microphysics,dynamics and precipitation.The cloud formation,intensityandevolutionas wellthe spatial andtemporaldistributionsofseededandnon-seededcloudswerepresented and compared.The study area was chosen around Beijing urban regions where the aerosol background number concentration was around 1026 cm-3,which was primarily produced by air pollutants and composed of small-sized particles.The seeded aerosol number concentration was 1000 cm-3with sizes larger than 1μm in diameter,which were referred to as giant particles in the study.The seeding duration was 10 min,beginning from 0730 LSTnd ending at 0740 LST.The seeding height was at 2.5-3.5 km,which was located just near the cloud base.The WRF model was initialized with NCEP data and run for 24 hours.

Comparing the non-seeded case with the seeded case,we found that the cloud seeding with water-soluble aerosols of sizes larger than 1.0μm in diameter may induce substantial changes in cloud microphysics,dynamics and precipitation.The indication was also that seeding has a high impact on cloud ice processes,such as increased ice particle concentrations.The inf uential time for seeding is around 3-4 hours.The maximum surface rain intensity,rain amount and the accumulated precipitation are greatly changed,in which the rain intensity is increased,but the maximum precipitation amount is decreased and surface accumulated precipitation is increased.The maximum increase of surface accumulated precipitation can be 41%.The surface precipitation distribution is also changed from the concentrated type for the non-seeded case to the more dispersed type for the seeded case(Fig.10).The precipitation distribution for the seeded caseindicatesanapparentprecipitationincreaseinthe“extraarea”and“downwind area”.

2.4.5.Scientif c basis for hail suppression

Cloud seeding for hail suppression is the second most frequent weather-modif cation activity in China.The scientif c basis for hail suppression in China follows two main mechanisms:one is based on the so-called benef cial competition mechanism by over-seeding of the ice nucleus,and the second is the explosion mechanism induced by artillery. Although the latter has not been scientif cally conf rmed in the weather-modif cationcommunity,the practical use of this method is still popular in China.

Researchers have developed an AgI-seeding process in a 2D hail cloud model to simulate the seeding effect on hail suppression(Huang and Xu,1994).The AgI-based nucleation processes such as deposition and contact freezing are included in the model to investigate the seeding effects on hail suppression with various hailstorms and seeding methods.The possible effect of explosion that causes the raindrop breakup on hail suppression has also been discussed(e.g., DuanandXu,2001)usinga 2D hailmodel.AgI-basednucleation processes have also been developed in a hail-bin model (Guo et al.,2003)to study the seeding effect on the hail suppression mechanism(Guo et al.,2004;Xiao et al.,2006). These studies show that,if appropriate seeding methods are used,the effect of seeding on hail clouds can signif cantly suppress hail formation by causing early and increased precipitation of rain.The application of hail-cloud models in the early identif cation of hail-cloud has also been conducted in China(Yu et al.,2001;Xiao et al.,2002;Li et al.,2003).

One of the key issues for operational hail suppression activities is to identify hailstorms and non-hail storms.Chinese scientists have developed identif cation criteria for this purpose based on f eld experiments(Table 1).The criteria for identifying strong hailstorms are quite good and the accuracy rate can reach more than 90%.

In recent years,dual-polarized radars have been used in hail-suppression operations.Figure 11 is an example of the Range Height Indicator(RHI)scan of a hail cloud.The location of hail formationis quite clear and can be a goodindexfor cloud-seeding operations.

Table 1.Identif cation criteria for hailstorms and non-hail storms in western China.

Although signif cant progress in the theoretical modeling of hailstorms,especially in modeling hail-cloud microphysics,hail formation mechanisms and the seeding effect, has been made in recent years in China,there are still many open questions that require clarif cation.Mesoscale cloud models with detailed hail microphysics and with realistic seeding routines should be applied during all stages of the forecasting,operation and evaluation of hail-suppression activities.The prediction ability of time,location and amount of hailfall in hailstorm models should be further improved. Conductingdetailed comparisonsbetween observedand simulated results of hailstorms is necessary in order to improve the physical processes in hail-cloud models.

3. Discussion and conclusions

Over the past several decades,great efforts and progress in cloud physics and weather modif cation have been made in China,including studies on cloud modeling,cloud microphysics and dynamics,aerosol-cloudinteractions,MCSs,severe wind formation,and weather modif cation.

More advanced cloud models with detailed microphysics and mesoscale cloud-resolving models have been developed and applied to guide cloud-seeding operations,as already done in many other countries outside China(e.g.,Tzivion et al.,1994;Bruintjes et al.,1995;Orville,1996).

The modern techniques of GPS,remote sensing and GIS havebeenwidelyusedinoperationalcloud-seedingprograms in China,andhave greatlydecreasedthe blindnessanduncertainties in weather-modif cation activities.In addition,cloud models with glaciogenic and hygroscopic cloud-seedingprocesses have been developed and used in order to understand the variations in cloud microphysics,dynamics and precipitation induced by cloud seeding.Furthermore,we have also seen a rapid increase in the interest in this area due to the strong demands for fresh water and a wide variety of disaster prevention programs.

Despite signif cant advances in relevant subjects of cloud physics and weather modif cation in recent years in China, there are still many def ciencies or challenges left to be addressed,such as gaining a better understanding of cloud and precipitation formation mechanisms,improving the seeding eff ciency and seeding techniques,and obtaining more reliable evidence of seeding effectiveness,as indicated in some other publications(National Research Council,2003;List, 2004;Garstang et al.,2005).The potential for progress in cloud physics and weather-modif cation research is dependent upon improvedcloud model treatments of cloud physics and precipitation,their resolutionand initialization via observational data and data assimilation techniques,and also upon applyingadvancedobservationaltechnologiesand accurately retrieving information from remote sensing data.

China is the largest country to conduct an operational weather-modif cation program in the world.Comparing the f eld with that outside China(National Research Council, 2003),the weather-modif cation activities in China have some unique features,one of which is the application of multiple advanced seeding tools in weather-modif cation activities,such as aircraft,ground-basedrockets,artillery and generators.These seeding tools can realize seeding operations in different cloud systems from stratiform clouds to severe convective clouds such as hailstorms.Rocket-based seeding candirectly and promptly deliver seeding material into the targeted area of severe clouds,which is impossible or diff cult foraircraftseedingduetosecurityreasons.Therefore,thedevelopment of reliable rocket and advanced cloud monitoring systems are critical to achieve successful seeding results.

Another feature is that operational weather-modif cation activities in China are mainly conducted to meet emergency needs in places that suffer severe drought and hail-related disasters,as compared to some countries who usually conduct cloud-seeding experiments in a f xed location.Therefore,to conduct long-term,well-designed and randomized cloud-seeding experiments is quite diff cult via the operational cloud-seeding program.The evaluation of seeding effectiveness has to depend on comparisons of rain gauge data betweentargetandcontrolareas,as wellas onphysicalevaluations of targeted clouds by radar or other observational data. To meet statistical requirements,the selection of a control area is often diff cult in places lacking climatic rain gauge data.

In the past ten years,we have witnessed rapid development of a multitude of advancedtools used for observing and seeding cloud systems in China.This progress provides an effective way to explore cloud structure and identify seeding targets,as well as determine effective seeding locations.It is becoming possible to conduct detailed studies on the chain of physical events involved in the effects induced by cloud seeding.Reducing key uncertainties in weather modif cation via long-term projects aimed at understandingcloud and precipitation microphysics and dynamics,as well as cloud modeling and cloud seeding evaluations are extremely urgent and important.An upcoming national project on weather modifcation in China will shed light on these issues.

Acknowledgements.The authors are grateful for the critical and valuable comments from the anonymous reviewers,which helped greatly to improve the quality of this paper.This research was jointly sponsored by the Chinese Natural Science Foundation (Grant Nos.41005072 and 40575003),the Key Science and Technology Supporting Project of the Ministry of Science and Technology of China(Grant Nos.2006BAC12B03 and GYHY200806001) and the Third Tibetan Plateau Scientif c Experiment:Observations for Boundary Layer and Troposphere(GYHY201406001)

REFERENCES

Atkins,N.T.,and M.S.Laurent,2009:Bow echo mesovortices. PartI:Processes that inf uence their damaging potential.Mon. Wea.Rev.,137,1497-1513.

Bornstein,R.,and Q.Lin,2000:Urban heat islands and summertime convective thunderstorms in Atlanta:Three cases studies.Atmos.Environ.,34,507-516.

Bruintjes,R.T.,1999:A review of cloud seeding experiments to enhance precipitation and some new prospects.Bull.Amer. Meteor.Soc.,80,805-820.

Bruintjes,R.T.,T.L.Clark,and W.D.Hall,1995:The dispersion of tracer plumes in mountainous regions in central Arizona: Comparisons between observations and modeling results.J. Appl.Meteorol.,34,971-88.

Changnon,S.A.,K.R.Gabriel,E.W.Nancy,and R.R.Czys, 1995:Exploratory analysis of seeding effects on rainfall:Illinois 1989.J.Appl.Meteor.,34,1215-1224.

Chen,B.J.,D.P.Zhou,F.J.Gong,J.H.Wang,and S.J.Geng, 2005:AgI-Seeding modeling study on the 12 July 2002 cold vortex precipitation in Shenyang.Journal of Nanjing Institute of Meteorology,28,483-491.(in Chinese)

Chen,L.,and Y.Yin,2008:Numerical simulations for the thermal impact onthelong-rang transport of mineral aerosols.PlateauMeteorology,27,628-636.(in Chinese)

Chen,L.,and Y.Yin,2009:A sensitivity study of the effect of dust aerosols on the development of ice-phase cloud processes.Scientia Meteorologica Sinica,29,208-213.(in Chinese)

Chen,L.,Y.Yin,J.Yang,and S.J.Niu,2007:Effects of sand dust particles on cloud and precipitation:A numerical study.Journal of Nanjing Institute of Meteorology,30,590-600.(in Chinese)

Chen,L.X.,X.J.Zhou,W.L.Li,Y.F.Luo,and W.Q.Zhu, 2004:Characteristics of the climate change and its formation mechanism in China in the last 80 years.Acta Meteorologica Sinica,62,634-645.(in Chinese)

Cheng,C.S.,1959:Experiment of rain enhancement in China.Acta Meteorologica Sinica,30,286-290.(in Chinese)

Cotton,W.R.,and A.Boulanger,1975:On the variability of“dynamic seedability”as a function of time and location over South Florida:Part I,Spatial variability.J.Appl.Meteor.,14, 710-718.

Cotton,W.R.,1986:Testing,implementation and evolution of seeding concepts—A review.Meteorol.Monogr.,21,63-70.

Czys,R.R.,S.A.Changnon,N.E.Westcott,R.W.Scott,and M. S.Petersen,1995:Responses of warm-based,midwest cumulus congestus to dynamic seeding trials.J.Appl.Meteor., 34,1194-1214.

Dai,J.,X.Yu,and D.Rosenfeld,2008:The suppression of aerosols to the orographic precipitation in the Qinling Mountains.Chinese J.Atmos.Sci.,32,1319-1332.(in Chinese)

Duan,J.,and J.T.Mao,2008:Progress in researches on interaction between aerosol and cloud.Adv.Earth Sci.,23,252-261. (in Chinese)

Duan,J.,and J.T.Mao,2009:Inf uence of aerosol on regional precipitation in North China.Chinese Sci.Bull.,54,474-483.

Duan,Y.,and H.B.Xu,2001:The possible cloud microphysical mechanism of explosion in hail suppression.Acta Meteorologica Sinica,59,334-340.(in Chinese)

Fan,S.X.,and X.L.An,2000:Measurement and analysis of the concentration of cloudcondensation Nuclei inMT.Helanshan Area.J.Des.Res.,20,338-340.(in Chinese)

Fan,Y.,X.L.Guo,D.H.Fu,and H.Y.Li,2007:Observational studies on aerosol distribution during August to September in 2004 over Beijing and its surround areas.Climatic Environ. Res.,12,49-62.(in Chinese)

Fan,Y.,X.L.Guo,D.G.Zhang,D.H.Fu,J.D.Chen,and Y.L. Ma,2010:Airborne particle measuring system measurement on structure and size distribution of stratocumulus during August to September in 2004 over Beijing and its surrounding areas.Chinese J.Atmos.Sci.,34,1187-1200.(in Chinese)

Fang,C.G,and X.L.Guo,2007:The physical and precipitation responses to AgI seeding from a WRF-based mesoscale seeding model.Proc.Ninth WMO Conference on Weather Modifcation,Antalya,Turkey,World Meteorological Organization.

Fang,C.G.,X.L.Guo,and P.X.Wang,2009:The physical and precipitation response to AgI-seeding from a mesoscale WRF-based seeding model.Chinese J.Atmos.Sci.,33,621-633.(in Chinese)

Fang,W.,2004:Study of convective clouds in Qinghai Province.Meteorological Science Technology,32,343-347.(in Chinese)

Fang,W.,2008:Effect of aerosol on cloud and precipitation.Ph. D.dissertation,Nanjing University of Information Science& Technology,China,158 pp.(in Chinese)

Fang,W.,G.G.Zheng,and Z.J.Hu,2002:Parameterization of physical processes for hailstone growth.Acta Meteorologica Sinica,60,698-705.(in Chinese)

Fang,W.,G.G.Zheng,X.L.Guo,and H.B.Xu,2003:Overview of hail suppression projects in China.Proc.of WMO Meeting of Experts on Hail Suppression,Nalchik,53-55.

Fang,W.,G.G.Zheng,and G.F.He,2005a:Simulation study on the precipitation and seeding process of convective-cloud in Qinghai with a 3D cloud model.Journal of Nanjing Institute of Meteorology,28,763-769.(in Chinese)

Fang,W.,G.G.Zheng,and Z.J.Hu,2005b:Numerical simulations of the physical process for hailstone growth.Acta Meteorologica Sinica,19,93-101.(in Chinese)

Fu,D.H.,and X.L.Guo,2006:A cloud-resolving study on the role of cumulus merger process in MCS heavy precipitation.Adv.Atmos.Sci.,23,857-868.

Fu,D.H.,and X.L.Guo,2007:Numerical study on a severe downburst-producing thunderstorm on 23 August 2001 in Beijing.Adv.Atmos.Sci.,24,227-238.

Fu,D.H.,and X.L.Guo,2012:A cloud-resolving simulation study on the merging processes and effects of topography and environmental winds.J.Atmos.Sci.,69,1232-1249,doi: 10.1175/JAS-D-11-049.1.

Fu,D.H.,X.L.Guo,W.A.Xiao,and L.F.Sun,2003:Numerical study on the formation a severe storm accompanied with gale and heavy rain in Beijing.Journal of Nanjing Institute of Meteorology,26,190-200.(in Chinese)

Fu,D.H.,X.L.Guo,and C.H.Liu,2011:Effects of cloud microphysics on monsoon convective system and its formation environments over the South China Sea:A twodimensional cloud-resolving modeling study.J.Geophys. Res.,116,D07108,doi:10.1029/2010JD014662.

Fukuta,N.,1996a:Project mountain valley sunshine progress in science and technology.J.Appl.Meteor.,35,1483-1493.

Fukuta,N.,1996b:Low level penetration seeding with homogeneous ice nucleant for optimization of the induced microphysics-dynamics interactions.Proc.13th Conf.on Planned and Inadvertant Weather Modif cation,Atlanta, 164-171.

Fukuta,N.,1998:Conceptual structure for optimal utilization of phase change energy in RETHIT-FILAS feedback seeding mechanism.Proc.of 14th Conf.on Planned and Inadvertant Weather Modif cation,Everett,WA,618-621.

Garstang,M.,R.Bruintjes,R.Seraf n,H.Orville,B.Boe,W.Cotton,and J.Warburton,2005:Weather modif cation:Finding common ground.Bull.Amer.Meteor.Soc.,86,647-655.

Garvey,D.M.,1975:Testing of cloud seeding materials at the cloud simulation and aerosol laboratory,1971-1973.J.Appl. Meteor.,14,883-890.

Grabowski,W.W.,2003:Impact of ice microphysics on multiscale organization of tropical convection in two-dimensional cloud resolving simulations.Quart.J.Roy.Meteor.Soc.,129, 67-81.

Gu,Z.C.,1962:Recent theoretical study on size-distribution of cloud and fog droplet.Acta Meteor.Sinica,32,267-284.(in Chinese)

Guo,X.,X.L.Guo,D.H.Fu,and S.J.Niu,2013:Relationship between bell-shaped terrain dynamic forcing,mountain wave propagation,and orographic clouds and precipitation.Chinese J.Atmos.Sci.,37,786-800,doi:10.3878/j.issn.1006-9895.2012.12114.(in Chinese)

Guo,X.L.,1997:The studies on three-dimensional hail-category numerical simulations of hail formation and hailfall pro-cesses.Ph.D.dissertation,Institute of Atmospheric Physics, Chinese Academy of Sciences,168 pp.(in Chinese)

Guo,X.L.,and M.Y.Huang,2002:Hail formation and growth in a 3D cloud model with hail-bin microphysics.Atmos.Res., 63,59-99.

Guo,X.L.,and D.H.Fu,2003:The formation process and cloud physical characteristics for a typical disastrous windproducing hailstorm in Beijing.Chinese Sci.Bull.,48(S2), 77-82.

Guo,X.L.,and G.G.Zheng,2009:Advances in weather modif cation from 1997 to 2007 in China.Adv.Atmos.Sci.,26, 240-252,doi:10.1007/s00376-009-0240-8.

Guo,X.L.,H.Niino,and R.Kimura,1999a:Numerical modeling on a hazardous microburst-producing hailstorm.Vol.1,Towards Digital Earth—Proc.of the International Symposium on Digital Earth,Beijing,Science Press,383-398.

Guo,X.L.,H.Niino,and R.Kimura,1999b:Numerical Simulation of a microburst-and hail-producing storm on 8 September 1994 in Saitama Prefecture.Proc.of 4th Symposium on Environmental Fluid Mechanics of Japan Soc.Fluid Mech., 4,501-502.

Guo,X.L.,M.Y.Huang,Y.C.Hong,H.Xiao,and H.C.Lei, 2000:A three-dimensional numerical simulation of structure and evolution of a multi-cellular hailstorm.Chinese J.Atmos. Sci.,24,51-58.(in Chinese)

Guo,X.L.,M.Y.Huang,Y.C.Hong,H.Xiao,and L.Zhou, 2001a:A study of three-dimensional hail-category hailstorm model Part I:Model description and the mechanism of hail recirculation growth.Chinese J.Atmos.Sci.,25,707-720.(in Chinese)

Guo,X.L.,M.Y.Huang,Y.C.Hong,H.Xiao,and L.Zhou, 2001b:A study of three-dimensional hail-category hailstorm model Part II:Characteristics of hail-category size distribution.Chinese J.Atmos.Sci.,25,395-404.(in Chinese)

Guo,X.L.,G.G.Zheng,and W.Fang,2003:Theoretical modeling of hailstorms in China.Proc.of WMO Meeting of Experts on Hail Suppression,Nalchik,63-65.

Guo,X.L.,H.Niino,and R.Kimura,2004:Modeling study of a microburst-producing storm on 8 September 1994 in Saitama prefecture.Ocean Monthly,38,162-167.

Guo,X.L.,G.G.Zheng,and D.Z.Jin,2006a:A numerical comparison study of cloud seeding by silver iodide and liquid carbon dioxide.Atmos.Res.,79,183-226.

Guo,X.L.,D.H.Fu,and J.Wang,2006b:Mesoscale convective precipitation system modif ed by urbanization inBeijing City.Atmos.Res.,82,112-126.

Guo,X.L.,D.H.Fu,and G.G.Zheng,2007:Modeling study on optimal convective cloud seeding in rain augmentation.Asia-Pacif c J.Atmos.Sci.,43,273-284.

Guo X.L.,D.H.Fu,and Z.X.Hu,2013:Progress incloud physics, precipitation,and weather modif cation during 2008-2012.Chinese J.Atmos.Sci.,37,351-363,doi:10.3878/j.issn. 1006-9895.2012.12321.(in Chinese)

Guo,X.L.D.H.Fu,X.Guo,and C.M.Zhang,2014:A case study of aerosol impacts on summer convective clouds and precipitation over northern China.Atmos.Res.,142,142-157,doi: 10.1016/j.atmosres.2013.10.006.

Hjelmfelt,M.R.,H.D.Orville,R.D.Roberts,J.P.Chen,and F. J.Kopp,1989:Observational and numerical study of a microburst line-producing storm.J.Atmos.Sci.,46,2731-2743.

Hong,S.-Y.,and J.-O.J.Lim,2006:The WRF single-moment 6-class microphysics scheme(WSM6).Journal of the KoreanMeteorological Society,42,129-151.

Hong,Y.C.,1998:A 3-D hail cloud numerical seeding model.Acta Meteorologica Sinica,56,641-653.(in Chinese)

Hong,Y.C.,1999:Study on mechanism of hail formation and hail suppression withseeding.Acta Meteorologica Sinica,57,30-44.(in Chinese)

Hong,Y.C.,and F.F.Zhou,2006:The study of evaluation of potential of artif cial precipitation enhancement in stratiform cloud systemChinese J.Atmos.Sci.,30,913-926.(in Chinese).

Hong,Y.C.,H.Xiao,H.Y.Li,and Z.X.Hu,2002:Studies on microphysical processes in hail cloud.Chinese J.Atmos.Sci., 26,421-432.(in Chinese)

Houze,R.A.,S.A.Rutledge.,and M.I.Biggerstaff,1989:Interpretation of Doppler weather radar displays of midlatitude mesoscale convective systems.Bull.Amer.Meteor.Soc.,70, 608-619.

Hu,Z.J.,and G.F.He,1987:Numerical simulation of microprocesses in cumulonimbus clouds(I):Microphysical model.Acta Meteorologica Sinica,45,467-485.(in Chinese)

Hu,Z.X.,H.Y.Li,H.Xiao,C.Y.Hong,M.Y.Huang,and Y.X. Wu,2003:Numerical simulation of hailstorms and the characteristics of accumulation zone of supercooled raindrops in Xunyi County.ClimaticandEnvironmental Research,8,196-208.(in Chinese)

Hu,Z.X.,H.C.Lei,X.L.Guo,D.Z.Jin,Y.B.Qi,and X.Q. Zhang,2007:Studies of the structure of a stratiform cloud and the physical processes of precipitation formation.Chinese J.Atmos.Sci.,31,425-439.(in Chinese)

Huang,J.P.,B.Lin,P.Minnis,T.Wang,X.Wang,Y.X.Hu,Y. H.Yi,and J.K.Ayers,2006a:Satellite-based assessment of possible dust aerosols semi-direct effect on cloud water path over EastAsia.Geophys.Res.Lett.,33,L19802,doi:10.1029/ 2006GL026561.

Huang,J.P.,P.Minnis,B.Lin,T.Wang,Y.H.Yi,Y.X.Hu,S.Sun-Mack,and K.Ayers,2006b:Possible inf uences of Asian dust aerosols on cloud properties and radiative forcing observed from MODIS and CERES.Geophys.Res.Lett.,33,L06824, doi:10.1029/2005GL024724.

Huang,M.Y.,and A.S.Wang,1980:Introduction of Hail Suppression.Science Press,204 pp.(in Chinese)

Huang,M.Y.,H.Y.Xu,and L.Zhou,2000:40 year’s hail suppression in China.Climatic and Environmental Research,5, 318-328.(in Chinese)

Huang,M.Y.,Z.L.Shen,and Y.C.Hong,2003:Advance of research on cloud and precipitation and weather modif cation in the latest half century.Chinese J.Atmos.Sci.,27,536-551. (in Chinese)

Huang,M.Y.,C.S.Zhao,G.Q.Zhou,Y.Duan,L.X.Shi,and Z.H.Wu,2005:Stratus cloud microphysical characters over North China region and the relationship between aerosols and clouds.Journal of Nanjing Institute of Meteorology,28,360-368.(in Chinese)

Huang,Y.,and H.Y.Xu,1994:Numerical experiments on hail suppression by AgI seeding.Scientia Atmospherica Sinica, 18,612-621.(in Chinese)

Jia,X.C.,and X.L.,Guo,2012a:Impacts of anthropogenic atmospheric pollutant on formation and development of a winter heavy fog event.Chinese J.Atmos.Sci.,36,995-1008.(in Chinese)

Jia,X.C.,and X.L.,Guo,2012b:Impacts of secondary aerosols on a persistent fog event in Northern China.Atmos.OceanicSci.Lett.,5,401-407.

Jin,D.Z,and Coauthors,2004:Measurement of column cloud liquid water content by airborne upward-looking microwave radiometer.Acta Meteorologica Sinica,62,868-874.(in Chinese)

Jin,H.,G.H.Wang,L.G.You,and D.X.Feng,2006:On physical structure of stratiform cloud during a precipitation process in Henan Province.Meteor.Mon.,32,3-10.(in Chinese)

Johnson,D.B.,1982:The role of giant and ultragiant aerosol particles in warm rain initiation.J.Atmos.Sci.,39,448-460.

Kang,F.Q.,Q.Zhang,S.P.Ma,Y.J.Qiao,and X.L.Guo, 2004a:Mechanism of hail formation on the northeast border of Qinghai-Xizang Plateau and its neighbourhood.Plateau Meteorology,23,749-757.(in Chinese)

Kang,F.Q.,Q.Zhang,Y.X.Qu,L.Z.Ji,and X.L.Guo,2004b: Simulating study on hail microphysical process on the northeastern side of Qinghai-Xizang plateau and its neighbourhood.Plateau Meteorology.,23,735-742.(in Chinese)

Kraus,E.B.,and P.A.Squires,1947:Experiments on the stimulation of clouds to produce rain.Nature,159,489-492.

Kong,F.Y.,1991:3D modeling on hail cloud.Ph.D.dissertation,Institute of Atmospheric Physics,Chinese Academy of Sciences,155 pp.(in Chinese)

Kong,F.Y.,M.Y.Huang,and H.Y.Xu,1990:A 3D simulation of ice phase process in convective clouds:model development and cold cloud parameterization.Chinese J.Atmos.Sci.,14, 441-453.(in Chinese)

Kong,F.Y.,M.Y.Huang and H.Y.Xu,1991:3D simulation of roles of ice phase in cumulus development.Science in China(B),15,78-88.

Lei,H.C.,C.Wei,Z.L.Shen,X.Q.Zhang,D.Z.Jin,S.F.Gu,M. L.Li,and J.H.Zhang,2003:Measurement of column cloud liquid water content by airborne upward-looking microwave radiometer I:Instrument and its calibration.Plateau Meteorology,22,551-557.(in Chinese)

Lei,H.C.,Y.C.Hong,Z.Zhao,H.Xiao,and X.L.Guo,2008:Advances in cloud and precipitation physics and weather modif cation in recent years.Chinese J.Atmos.Sci.,32,967-974. (in Chinese)

Li,H.Y.,Z.X.Hu,H.,Xiao,and Y.C.Hong,2003:Numerical studies of the practical seeding methods in hail suppression.Chinese J.Atmos.Sci.,27,212-222.(in Chinese)

Li,H.Y.,H.Wang,and Y.C.Hong,2006:A numerical study of precipitation enhancement potential in frontal cloud system.Chinese J.Atmos.Sci.,30,341-350.(in Chinese)

Li,J.,and J.T.Mao,2006:Inf uence of atmospheric ice nucleus concentrations oncoldcloudradiant properties and coldcloud ref ectivity changes in past years.Chinese Science Bulletin, 51,480-489.

Li,J.,R.C.Yu,T.J.Zhou,and B.Wang,2005:Why is there an early spring cooling shift downstream of the Tibetan Plateau?J.Climate,18,4660-4668.

Li,X.S.,X.L.Guo,D.H.Fu,and B.Yang,2011:Studies of evolution features and persistent development mechanism of mesoscale convective clouds over the Northern South China Sea.Chinese J.Atmos.Sci.,35,259-271,doi:10.3878/j.issn. 1006-9895.2011.02.06.(in Chinese)

Li,X.S.,X.L.Guo and D.H.Fu,2013:TRMM-retrieved cloud structure and evolution of MCSs over the Northern South China Sea and impacts of CAPE and vertical wind shear.Adv. Atmos.Sci.,30,77-88,doi:10.1007/s00376-012-2055-2.

Li,X.Y.,X.L.Guo,and J.Zhu,2008:Climatic features of cloud water distribution and cycle over China.Adv.Atmos.Sci.,25, 437-446,doi:10.1007/s00376-008-0437-2.

Li,Y.Y.,R.C.Yu.,Y.P.Xu.,and X.H.,Zhang,2004:Spatial distribution and seasonal variation of cloud over China based on ISCCP data and surface observations.J.Meteor.Soc.Japan, 82,761-773.

Lin,W.S.,S.J.Fan,X.M.Wang,M.Y.Huang,L.Zhou,and X.L.Guo,2000:Numerical simulation of the hail cloud over mountainous terrain.PlateauMeteor.,19,59-65.(inChinese)

List,R.,2004:Weather modif cation—A scenario for the future,Bull.Amer.Meteor.Soc.,85,51-63.

Liu,J.,M.L.Li,T.Jiang,C.S.Mi,and Z.X.Chen,2005:The preliminarystudyof thebasicstructureof precipitatingstratus and precipitation potential in spring in JilinProvince.Scientia Meteorologica Sinica,25,609-616.(in Chinese)

Liu,W.G.,Z.J.Su,G.H.Wang,and S.R.Li,2003:Development and application of new-generation airborne particle measuring system.J.Appl.Meteor.Sci.,14(S1),11-18.(in Chinese)

Liu,X.E.,and X.L.Guo,2012a:Analysis and numerical simulation researches on surface damaged wind formation mechanism and structural characteristics of a squall line case.Chinese J.Atmos.Sci.,36(6),1150-1164,doi:10.3878/j.issn. 1006-9895.2012.11212.(in Chinese)

Liu,X.E.,and X.L.Guo,2012b:Role of downward momentumtransport.intheformation of severe surface winds.Atmos. Oceanic Sci.Lett.,5,379-383.

Lou,X.F.,Z.J.Hu,P.Y.Wang,and X.J.Zhou,2003:Introduction to microphysical scheme of mesoscale atmospheric models and cloud models.J.Appl.Meteor.Sci.,14(S1),49-59.(in Chinese)

Lou,X.F.,Y.Q.Shi,J.Sun,L.L.Xue,Z.J.Hu,W.Fang,and W. G.Liu,2012:Cloud-resolving model for weather modif cation in China.Chinese Science Bulletin,57,1055-1061,doi: 10.1007/s11434-011-4934-9.

Lu,G.X.,and X.L.Guo,2012:Distribution and origin of aerosol and its transform relationship with CCN derived from the spring multi-aircraft measurements of Beijing Cloud Experiment(BCE).Chinese Science Bulletin,57,2460-2469,doi: 10.1007/s11434-012-5136-9.

L¨u,J.M.,J.Z.Ren,and J.H.Ju,2004:Theinterdecadal variability of East Asia monsoon and its effect on the rainfall over China.Journal of Tropical Meteorology,10,73-80.(in Chinese)

Ma,J.Z.,X.L.Guo,C.S.Zhao,Y.J.Zhang,and Z.J.Hu,2007: Recent progress in cloud physics research in China.Adv.Atmos.Sci.,24,1121-1137,doi:10.1007/s00376-007-1121-7.

Malkus,J.S.,and G.Witt,1959:The evolution of a convective element:A numerical calculation.The Atmosphere and Sea in Motion,Rockefeller Institute,Oxford University Press,425-439.

Mao,J.T.,and G.G.Zheng,2006:Discussions on some weather modif cation issues.J.Appl.Meteor.Sci.,17,643-646.(in Chinese)

Mao,J.T.,P.X.Pei,and Z.F.Tian,1982:An approach of the hail forecasting with a numerical model of cumulus.Acta Meteorologica Sinica,40,73-79.(in Chinese)

National Research Council,2003:Critical Issues in Weather Modif cation Research.National Academy Press,Washington, DC.,123 pp.

Orville,H.D.,1996:A review of cloud modeling in weather modif cation.Bull.Amer.Meteor.Soc.,77,1535-1555.

Orville,H.D.,and F.L.Kopp,1977:Numerical simulation of the life history of a hailstorm.J.Atmos.Sci.,34,1596-1618.

Orville,H.D.,and J.M.Chen,1982:Effects of cloud seeding,latent heat of fusion,andcondensate loading oncloud dynamics and precipitation evolution:A numerical study.J.Atmos.Sci., 39,2807-2827.

Pruppacher,H.R.,and J.D.Klett,1997:The Microphysics of Clouds and Precipitation.2d ed.,D.Reidel,714 pp.

Qi,L.B.,M.Y.Huang,H.Xiao,and Y.C.Hong,2002a:A numerical study of hail identif cation using multi-wavelength radar.Chinese J.Atmos.Sci.,26,30-40.(in Chinese)

Qi,L.B.,M.Y.Huang,H.Xiao,and Y.C.Hong,2002b:A numerical study of hail identif cation using dual-polarization radar.Chinese J.Atmos.Sci.,26,230-240.(in Chinese)

Qi,Y.B.,X.L.Guo,and D.Z.Jin,2007:An observational study of macro/microphysical structures of convective rain bands of a cold vortex over northeast China.Chinese J.Atmos.Sci.,31, 621-634.(in Chinese)

Rogers,D.C.,and P.J.DeMott,1991:Advances in laboratory cloud physics 1987-1990.Rev.of Geophys.,29(S1),80-87.

Rosenfeld,D.,and W.L.Woodley,1993:Effects of cloud seeding in West Texas:Additional results and new insights.J.Appl. Meteor.,32,1848-1866.

Rosenfeld,D.,X.Yu,and J.Dai,2005:Satellite-retrieved microstructure of AgI seeding tracks in supercooled layer clouds.J.Appl.Meteor.,44,760-767.

Rozoff,C.,W.R.Cotton,and J.O.Adegoke,2003:Simulation of St.Louis,Missouri,land use impacts on thunderstorms.J. Appl.Meteor.,42,716-738.

Saunders,P.M.,1957:The thermodynamics of saturated air:A contribution to the classical theory.Quart.J.Roy.Meteor. Soc.,83,342-350.

Schaefer,V.J.,1946:The production of ice crystals in a cloud of supercooled water droplets.Science,104,457-459.

Shepherd,J.M.,H.Pierce,and A.J.Negri,2002:Rainfall modif cation by major urban areas:Observations from spaceborne rain radar on the TRMM satellite.J.Appl.Meteor.,41,689-701.

Shi,L.X.,and Y.Duan,2007:Observations of cloud condensation nuclei in north China.Acta Meteorologica Sinica,65, 644-652.(in Chinese)

Silverman,B.A.,2000:An independent statistical reevaluation of the South African hygroscopic f are seeding experiment.J. Appl.Meteor.,39,1373-1378.

Silverman,B.A.,2001:A critical assessment of glaceogenic seeding of convective clouds for rainfall enhancement.Bull.Amer. Meteor.Soc.,82,903-921.

Silverman,B.A.,and W.Sukarnjanasat,2000:Results of the Thailand warm-cloud hygroscopic particle seeding experiment.J. Appl.Meteor.,39,1160-1175.

Su,Z.J.,W.G.Liu,G.H.Wang,C.H.Xu,and L.J.Wang, 2003a:Microphysical characteristics of a precipitation process in Qinghai Province.J.Appl.Meteor.Sci.,14(S1),27-35.(in Chinese)

Su,Z.J.,G.H.Wang,W.G.Liu,and L.G.Li,2003b:Microstructure analysis of spring precipitable clouds in Qinghai Province.J.Appl.Meteor.Sci.,14(Suppl.),36-40.(in Chinese)

Sulakvelidze,G.K.,N.S.Bibilashvei and V.F.Lapcheva,1967:Formation of Precipitation,and Modif cation of Hail Processes.Israel Program for Scientif c Translations,208 pp.

Sun,L.F.,X.L.Guo,L.T.Sun,and D.H.Fu,2004:A numerical study of an airplane disaster-producing microburst on 22 June 2000 in Wuhan.Chinese J.Atmos.Sci.,27,307-323.(in Chinese)

Takahashi,T.,1976:Warm rain,Giant Nuclei and Chemical balance—A numerical model.J.Atmos.Sci.,33,269-286.

Takahashi,T.,and S.Lee,1978:The nuclei mass range most effcient for the initiation of warm cloud showers.J.Atmos.Sci., 35,1934-1946.

Tao,W.-K.,2007:Cloud resolving modeling.J.Meteor.Soc. Japan,85,305-330.

Tao,W.-K.,and M.W.Moncrieff,2009:Multi-scalecloud-system modeling.Rev.Geophys.,47,RG4002,doi:10.1029/2008RG 000276.

Tao,W.-K.,C.-L.Shie,J.Simpson,S.Braun,R.H.Johnson, and P.E.Ciesielski,2003:Convective systems over the South China Sea:Cloud-resolving model simulations.J.Atmos. Sci.,60,2929-2956.

Trapp,R.J.,and M.L.Weisman,2003:Low-level mesovortices within squall lines and bow echoes.Part II:Their genesis and implications.Mon.Wea.Rev.,131,2804-2823.

Tzivion,S.,S.Reisin,and Z.Levin,1994:Numerical simulation of hygroscopic seeding in convective clouds.J.Appl.Meteor., 33,252-267.

Trenberth,K.E.,J.Fasullo,and L.Smith,2005:Trends and variability in column-integrated atmospheric water vapor.Climate Dyn.,24,741-758.

Vonnegut,B.,1947:The nucleation of ice formation by silver iodide.J.Appl.Phys.,18,593-595.

Wang,H.,H.C.Lei,Deligeer,G.L.Li,and W.A.Xiao,Y.C. Hong,M.Y.Huang,2002:A numerical simulation of characteristics of convective cloud at the upper reaches of the yellow river.Climatic and Environmental Research,7,397-408.(in Chinese)

Wang,H.J.,2001:The weakening of the Asian monsoon circulation after the end of 1970’s.Adv.Atmos.Sci.,18,376-385.

Wang,Y.F.,H.C.Lei,Y.X.Wu,W.A.Xiao,and X.Q.Zhang, 2005:Size distributions of the water drops in the warm layer of stratiform clouds in Yan’an.Journal of Nanjing Institute of Meteorology,28,787-793.(in Chinese)

Wang,Y.L.,and H.C.Lei,2003:Test of cold cloud seeding.Chinese J.Atmos.Sci.,27,929-938.(in Chinese)

Weisman,M.L.,and J.B.Klemp,1982:The dependence of numerically simulated convective storms on vertical wind shear and buoyancy.Mon.Wea.Rev.,110,504-520.

Weisman,M.L.,and J.B.Klemp,1984:The structure and classif cation of numerically simulated convective storms in directionally varying wind shear.Mon.Wea.Rev.,112,2479-98.

WMO(World Meteorological Organization),2000:Report of the WMO International Workshop on Hygroscopic Seeding:Experimental Results,Physical Processes and Research Needs. WMO,Geneva,68 pp.

Xiao,H.,H.Y.Xu,and M.Y.Huang,1988:The numerical simulation study of drops distributor in cumulus.Part I:The effect of salt nuclei spectrum and density.Scientia Atmospherica Sinica,12,121-130.(in Chinese)

Xiao,H.,Y.X.Wu,Z.X.Hu,H.Y.Li,H.C.Lei,Y.C.Hong,M. Y.Huang,P.Fan,G.Liang,T.M.Tie,2002:Earlier identif cation and numerical simulation of hail storms occurring in Xunyi region.Plateau Meteor.,21,159-166.(in Chinese)

Xiao,M.J,X.L Guo,and W.A.Xiao,2006:Numerical modeling study on hail suppression and rain enhancement by seeding convective clouds with silver iodide and liquid carbon dioxide.Journal of Nanjing Institute of Meteorology,29,48-55. (in Chinese)

Xu,H.B.,and S.W.Wang,1985:A numerical model of hailbearing convective cloud I:Bi-parameter evolution of size distribution of raindrops,frozen raindrops and hailstones.Acta Meteorologica Sinica,43,13-25.(in Chinese)

Xu,H.B.,and S.W.Wang,1988:Two dimension hail cloud model.Acta Meteorologica Sinica,46,227-236.(in Chinese)

Xu,H.B.,and Y.Duan,2001:The mechanism of hailstone’s formation and the hail suppression hypothesis:“benef cial competition”.Chinese J.Atmos.Sci.,25,277-288.(in Chinese)

Xu,H.B.,and Y.Duan,2002:The accumulation of hydrometeor and depletion of cloud water in strongly convective cloud (hailstorm).Acta Meteorologica Sinica,60,575-584.(in Chinese)

Xu,H.Y.,and Z.C.Gu,1963:Warm and thin cloud precipitation formed under f uctuation collision condition.Acta Meteorologica Sinica,33,108-114.(in Chinese)

Xu,H.Y.,and G.H.Xia,1964:The hail formation under the condition of f uctuant vertical velocities.Introduction of Hail Suppression,M.Y.Huang and A.S.Wang,Eds.,1980,Science Press,89-90.(in Chinese)

Xu,H.Y.,P.Q.Huang,M.Y.Huang,and J.P.Hao,1983:A study on the growth of cloud droplets by condensation in cumulus clouds.Chinese J.Atmos.Sci.,7,249-259.(in Chinese)

Xu,J.L.,1985:A numerical experiment on the hail growth in a given three-dimensional cloud model.Acta Meteor.Sinica, 43,469-478.(in Chinese)

Yang,W.X.,S.J.Niu,J.G.Wei,and Y.W.Sun,2005:Inhomogeneity of stratiform clouds in returning weather processes.Meteorological Science and Technology,33,256-259. (in Chinese)

Yao,Z.Y.,2006:Review of weather modif cation research in Chinese Academy of Meteorological Sciences.J.Appl.Meteor. Sci.,17,786-795.(in Chinese)

Yin,Y.,and L.Chen,2007:The effects of heating by transported dust layers on cloud and precipitation:A numerical study.Atmos.Chem.Phys.,7,3497-3505.

Yin,Y.,Z.Levin,T.G.Reisin,and S.Tzivion,2000:The effect of giant cloud condensation nuclei on the development of precipitation in convective clouds:A numerical study.Atmos. Res.,53,91-116.

Yin,Y.,S.Wurzler,Z.Levin,and T.G.Reisin,2002:Interactions of mineral dust particles and clouds:Effects on precipitation and cloud optical properties.J.Geophys.Res.,107,AAC 19-1-AAC 19-14,doi:10.1029/2001JD001544.

You,L.,P.Ma,and Z.Hu,2002b:A study on experiment of artifcial precipitation in North China.Meteorological Science and Technology,30(S1),19-56.(in Chinese)

You,L.G.,S.Z.Yang,X.G.Wang,and J.X.Pi,2002a:Study of ice nuclei concentration at Beijing in spring of 1995 and 1996.Acta Meteor.Sinica,60,101-109.(in Chinese)

Yu,D.W.,G.F.He,Y.Zhou,X.Y.Huang,and J.Z.Hu,2001: Three-dimensional convective cloud seeding model and its f eld application.Quart.J.Appl.Meteor.,12,122-132.(in Chinese)

Yu,R.C.,B.Wang,and T.J.Zhou,2004:Climate effects of the deep continental stratus clouds generated by Tibetan Plateau.J.Climate,17,2702-2713.

Yu,X.,J.Dai,D.Rosenfeld,H.Lei,X.Xu,P.Fan,and Z.Chen, 2005:Comparison of model-predicted transport and diffusion of seeding material with NOAA satellite-observed seeding track in supercooled layer clouds.J.Appl.Meteor.,44, 749-759.

Zhai,P.,and R.E.Eskridge,1997:Atmospheric water vapor over China.J.Climate,10,2643-2652.

Zhang,D.G.,X.L.Guo,D.H.Fu,and H.Y.Li,2007:Aircraft observation on cloud microphysics in Beijing and its surrounding regions during August-September 2003.Chinese J.Atmos.Sci.,31,596-610.(in Chinese)

Zhang,D.G.,X.L.Guo,D.L.Gong,and Z.Y.Yao,2011:The observational results of theclouds microphysical structure based on the data obtained by 23 sorties between 1989 and 2008 in Shandong Province.Acta Meteorologica Sinica,69,195-207. (in Chinese)

Zhang,L.M.,D.V.,Michelangeli,and P.A.Taylor,2006:Inf uence of aerosol concentration on precipitation formation in low-level,warm stratiform clouds.Journal of Aerosol Science,37(2),203-217.

Zhang,Q.,C.S.Zhao,X.X.Tie,Q.Wei,G.H.Li,and C.C.Li, 2006:Characterizations of Aerosols over the Beijing Region: A case study of aircraft measurements.Atmos.Environ.,40, 4513-4527.

Zhao,C.S.,D.Z.Zhang,and Y.Qin,1998:Simulation of the formation and development of aerosol in marine boundary layer.Progress in Natural Science,8,440-448.(in Chinese)

Zhao,C.S.,Y.Ishizaka,and D.Y.Peng,2005a:Numerical study on impacts of multi-component aerosols on marine cloud microphysical properties.J.Meteor.Soc.Japan,83,977-986.

Zhao,C.S.,D.Peng,and Y.Duan,2005b:The impacts of sea-salt and nss-sulfate aerosols on cloud microproperties.J.Appl. Meteor.Sci.,16,417-425.(in Chinese)

Zhao,C.S.,and Coauthors,2006a:Aircraft measurements of cloud droplet spectral dispersion and implications for indirect aerosol radiative forcing.Geophys.Res.Lett.,33,L16809, doi:10.1029/2006GL026653.

Zhao,C.S.,X.X.Tie,and Y.P.Lin,2006b:A possible positive feedback of reduction of precipitationand increase inaerosols over eastern central China.Geophys.Res.Lett.,33,L11814, doi:10.1029/2006GL025959.

Zheng,G.G.,X.L.Guo,W.Fang,and H.B.Xu 2003:Overview of hail suppression methods in China.Proc.of WMO Meeting of Experts on Hail Suppression,Nalchik,15-17.

Zhou,D.P.,F.J.Gong,J.H.Wang,Z.H.Li,and J.C.Gao,2004: The studies of the microphysical characteristics of aircraft cloud seeding.Scientia Meteorologica Sinica,24,405-412. (in Chinese)

Zhou,L.,B.J.Chen,Z.H.Li,M.Y.Huang,X.L.Guo,H.Xiao, and Y.C.Hong,2001:A numerical simulation of hailstorm accumulation zone and hail formation.Chinese Journal of Atmospheric Science,25,536-550.(in Chinese)

Zhou,Y.Q.,B.J.,Chen,H.Xiao,M.Y.Huang,and Z.H.Li, 2003:A case study of hail suppression by AgI seeding using 3D hailstorm model.Chinese J.Atmos.Sci.,27,8-22.(in Chinese)

:Guo,X.L.,D.H.Fu,X.Y.Li,Z.X.Hu,H.C.Lei,H.Xiao,and Y.C.Hong,2015:Advances in cloud physics and weather modif cation in China.Adv.Atmos.Sci.,32(2),230-249,

10.1007/s00376-014-0006-9.

(Received 15 May 2014;revised 1 August 2014;accepted 6 September 2014)

?Corresponding author:GUO Xueliang

Email:guoxl@mail.iap.ac.cn

Well-equipped aircraft and ground-based advanced Doppler and polarized radars have been rapidly applied in cloudseeding operations.The combined use of modern techniques such as the Global Positioning System,remote sensing,and Geographical Information Systems has greatly decreased the blindness and uncertainties in weather-modif cation activities. Weather-modif cation models based on state-of-the-art cloud-resolving models are operationally run at the National Weather Modif cation Centre in China for guiding weather-modif cation programs.

Despite important progress having been made,many critical issues or challenges remain to be solved,or require stronger scientif c evidence and support,such as the chain of physical events involved in the effects induced by cloud seeding.Current important progresses in measurements and seeding techniques provide the opportunity and possibility to reduce these def ciencies.Long-term scientif c projects aimed at reducing these key uncertainties are extremely urgent and important for weather-modif cation activities in China.