Role of Water Vapor Content in the Effects of Aerosol on the Electriﬁcation of Thunderstorms: a Numerical Study

atmosphere

Article Role of Water Vapor Content in the Effects of Aerosol on the Electriﬁcation of Thunderstorms: A Numerical Study

Pengguo Zhao 1,*, Yan Yin 2, Hui Xiao 3, Yunjun Zhou 1,4 and Jia Liu 5,6

1 Plateau Atmosphere and Environment Key Laboratory of Sichuan Province, College of Atmospheric Science, Chengdu University of Information Technology, Chengdu 610225, China; [email protected] 2 Key Laboratory for Aerosol-Cloud-Precipitation of China Meteorological Administration, Nanjing University of Information Science & Technology, Nanjing 210044, China; [email protected] 3 Institute of Tropical and Marine Meteorology, CMA, Guangzhou 510080, China; [email protected] 4 Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters, Nanjing University of Information Science & Technology, Nanjing 210044, China 5 Heavy Rain and Drought-Flood Disasters in Plateau and Basin Key Laboratory of Sichuan Province, Chengdu 610072, China; [email protected] 6 Climate Center of Sichuan Province, Chengdu 610072, China * Correspondence: [email protected] or [email protected]; Tel.: +86-28-8596-6389

Academic Editor: Robert W. Talbot Received: 4 September 2016; Accepted: 16 October 2016; Published: 20 October 2016

Abstract: We explored the role of the water vapor content below the freezing level in the response of idealized supercell storm electrical processes to increased concentrations of cloud condensation nuclei (CCN). Using the Weather Research and Forecasting model coupled with parameterizations electrification and discharging, we performed 30 simulations by varying both the CCN concentration and water vapor content below the freezing level. The sensitivity simulations showed a distinct response to increased concentrations of CCN, depending on the water vapor content below the freezing level. Enhancing CCN concentrations increased electrification processes of thunderstorms and produced a new negative charge region above the main positive charge center when there were ample amounts of water vapor below the freezing level. Conversely, there were weak effects on electrification and the charge structure in numerical experiments initialized with lower water vapor content below the freezing level.

Keywords: aerosol effects; water vapor content; electrical activity; thunderstorm; WRF model

1. Introduction Previous simulation studies on the effect of aerosol particles on thunderstorm electrification [1–3] have shown that increasing the aerosol concentration leads to the formation of numerous small droplets. These small droplets suppress the collision-coalescence processes and the formation of raindrops. They are then transported above the freezing level, where they enhance the development of ice particles. More ice particles participate in the electrification process under polluted conditions, resulting in an enhancement of this process. Observational studies [4–9] have indicated that anthropogenic aerosol particles produced from metropolitan regions, highways, and factories increase the intensity of thunderstorms, and the density and frequency of lightning. The emission of smoke from forest fires also has a marked impact on the electric processes of thunderstorms, leading to an increase in lightning activity around and downwind of forest regions, an increase in the percentage of positive cloud-to-ground lighting flashes, and a reduction in the percentage of negative cloud-to-ground lighting flashes [10–13].

Atmosphere 2016, 7, 137; doi:10.3390/atmos7100137 www.mdpi.com/journal/atmosphere Atmosphere 2016, 7, 137 2 of 19

Aerosol particles influence the electrification processes in thunderstorms by influencing liquid water and ice-phase particles. However, the effect of aerosol particles on the microphysical processes and precipitation of convective clouds is different under different environmental conditions. Saleeby et al. [14–16] investigated winter snowfall events over the Park Range of Colorado using Colorado State University’s Regional Atmospheric Modeling System. Their results showed that an increase in aerosol particles leads to a reduction in the size of cloud droplets and an increase in the concentration of droplets, although the total amount of precipitation was not changed. Tao et al. [17] showed that the effects of aerosols on convective systems, including their microphysical processes and precipitation, changed under different environmental conditions, by discussing a case of sea breeze convection, a tropical mesoscale convective system, and a mid-latitude squall line. Relative humidity plays a crucial part in the effect of aerosol particles on clouds [18]. Under conditions of low relative humidity, aerosol particles have little impact on the microphysical processes of clouds and the amount of precipitation, whereas they have an obvious impact on clouds under conditions of high relative humidity. Stevens and Feingold [19] showed that clouds can be interpreted as a buffered system in which significant changes in aerosol concentration have little effect on surface precipitation as a result of the complicated dynamic and microphysical processes in clouds. Saleeby et al. [16] showed that the influence of aerosol particles on convective clouds is important under the conditions of high relative humidity and large riming growth, whereas it is unremarkable when the amount of supercooled water is low. To investigate the aerosol effect on clouds under different conditions of low-level water vapor, Carrió and Cotton [20] showed that an increase in aerosol particles increases the content of supercooled water and enhances the growth of riming and ice-phase precipitation under conditions with a high amount of low-level moisture. Increasing aerosol concentration had little impact on microphysical processes and ice-phase precipitation under conditions of a low amount of low-level moisture. As the concentration of aerosol particles increases, the amount of supercooled water and ice-phase particles may also increase, leading to more ice-phase particles participating in the electrification process and enhancing the intensity of lighting activity during thunderstorms. Under conditions of a high moisture content and obvious riming growth, aerosol particles affect the microphysical processes of the cloud, but show little effect at low moisture content. The purpose of our study is to investigate how the effect of aerosol particles on thunderstorm electrification processes is sensitive to changes in environmental conditions, such as the amount of water vapor. The effect of aerosol particles on electrical activity may be different under different water vapor conditions, because the electrical activity is mainly influenced by ice-phase microphysical processes. To discuss the effect of aerosols on the electrification of thunderstorms at different water vapor content, the Weather Research and Forecasting (WRF) model, coupled with electrification and discharge parameterizations, was used to simulate electrification for different amounts of water vapor below the freezing level and different concentrations of initial cloud condensation nuclei (CCN).

2. Brief Description of the Model There is a general consensus that the main electriﬁcation mechanism in thunderstorm is non-inductive. Non-inductive electriﬁcation and a simple discharge scheme are coupled in the Morrison two-moment bulk mixed-phase microphysics scheme [21,22] of the WRF (v 3.4.1) model for discussing the effects of vapor condition under the freezing level and different initial CCN concentrations on the electrical development of thunderstorm. The Morrison microphysics scheme forecasts the mixing ratios and the concentration of cloud droplets, raindrops, ice crystals, snow particles, and graupel particles. Detailed microphysical processes are considered in the Morrison scheme, such as condensation and auto-conversion of cloud droplets; sublimation, deposition, and accretion of ice crystals; and ice multiplication, heterogeneous, and homogeneous freezing of raindrops and cloud droplets. Atmosphere 2016, 7, 137 3 of 19

Non-inductive electrification is considered in this paper, resulting from rebounding collisions between large and small ice particles. The charge separation rate between large and small ice particles is [23]: ∂ρ = β∆Q 1 − E E −1 n (1) ∂t xy xy xacy where ∆Q is the mean rebounding collision charge separation amount, Exy is the collection efficiency, nxacy is the collection rate between large and small ice particles, and β is the limited coefficient for charge separation rate. The mean rebounding collision charge separation amount in SP98 scheme is:

a b ∆Q = Bd V δq± (2) where B, a, and b are constants [23]; V is the relative fall speed; and d is the diameter of ice crystal or snow particle. The charge amount and polarity of a graupel particle (δq) are a function of the rime accretion rate RAR and the critical rime accretion rate RARC, if RAR > RARC, then the graupel are negatively charged, and vice versa. The electric potential Φ is computed through solving the Poisson equation: ρ ∇2Φ = − (3) ε

The electric permittivity ε is constant (8.859 × 10−12 N−1·m−2·C2); ρ is the charge density at the grid point. The electric ﬁeld E is solved through calculating the electric potential negative difference [24]: E = −∇Φ (4)

Further details of the microphysical, electriﬁcation, and discharge processes in this model have been discussed by Zhao et al. [3].

3. Experimental Design This study conducted idealized three-dimensional (3D) simulations of thunderstorm clouds. The simulation domain was 160 km × 80 km × 20 km, with a horizontal resolution of 0.5 km and 41 vertical levels. The total simulation time was 60 min and the time step was 2 s. The initial thermodynamic sounding profile (Figure1) was based on that of Weisman and Klemp [ 25] and was used to trigger an ideal continental thunderstorm, which was used as the baseline sounding. A very typical supercell thunderstorm can be simulated by using this thermodynamic sounding profile. Therefore many previous studies [23,26–30] discussed the cloud microphysics, precipitation, and electrical activities of an ideal continental thunderstorm by using the thermodynamic sounding profile used in our paper. These studies prove that the macro- and micro-physical structure and electrical characteristics of a typical continental thunderstorm can be simulated by using this thermodynamic sounding profile. To explore sensitivity tests under different water vapor conditions and initial CCN concentrations, these parameters were varied simultaneously. In this study, we assumed that the aerosol particles consist of ammonium sulfate with 100% solubility, and serve as CCN to influence the microphysical and electrical processes of a thunderstorm. The concentration of CCN exponentially decreases with height. Thirty numerical experiments were conducted by varying the water vapor conditions and CCN concentrations. Figure2 illustrates the sensitivity experiments conducted in this study. The initial CCN concentration was changed from 300 to 1500 cm–3 at 300 cm–3 intervals to represent changes in the concentration of aerosol particles. The sounding profile of Weisman and Klemp [25] was used to change the vapor mixing ratio (qv) below the freezing level to represent the variation in the water vapor conditions [20]. The water vapor mixing ratio was decreased to 85% in intervals of 3% with Atmosphere 2016, 7, 137 4 of 19 respect to the baseline sounding (without changing the water vapor content). The water vapor mixing ratiosAt ofmosphere the sensitivity 2016, 7,137 tests were about 97%, 94%, 91%, 88%, and 85% of the baseline sounding.4 of 19

Atmosphere 2016, 7,137 4 of 19

FigureFigure 1. The 1. The baseline baseline thermodynamic thermodynamic sounding profile proﬁle used used in inthe the simulation. simulation. Figure 1. The baseline thermodynamic sounding profile used in the simulation.

Figure 2. Each point represents an individual simulation. The horizontal axis denotes simulations with increasing initial CCN concentrations. The vertical axis shows the water vapor content with Figure 2. Each point represents an individual simulation. The horizontal axis denotes simulations Figurerespect 2. Each to the point sounding. represents an individual simulation. The horizontal axis denotes simulations with with increasing initial CCN concentrations. The vertical axis shows the water vapor content with increasing initial CCN concentrations. The vertical axis shows the water vapor content with respect to respect to the sounding. the4. Results sounding. and Discussion 4. Results and Discussion 4.1. Aerosol Effects on the Intensity of Electrical Activity and Sensitivity to Vapor Conditions 4. Results and Discussion 4.1. AerosolFigure Effects 3 shows on the the Intensity space of–time Electrical peak Activity values andof the Sensitivity positive to and Vapor negati Conditionsve charge densities at 4.1. Aerosol18 min Effectsof simulation, on the Intensity i.e. before of Electricalthe first flash Activity when and the Sensitivity positive toand Vapor negative Conditions charges had not Figure 3 shows the space–time peak values of the positive and negative charge densities at yet been neutralized in the thunderstorm. The peak values of the positive and negative charge 18Figure min of3 simulation,shows the i.e. space–time before the peakfirst flash values when of thethe positive and and negative negative charge charges had densities not at densities intuitively reflect the intensity of electrification intensity in the thunderstorm. In the cases yet been neutralized in the thunderstorm. The peak values of the positive and negative charge 18 minwith of simulation,baseline sounding i.e., before (without the changingfirst flash the when water the vapor positive condition), and negative the peak charges value had of the not yet densities intuitively reflect the intensity of electrification intensity in the thunderstorm. In the cases beenpositive neutralized charge in the density thunderstorm. increased significantly The peak values from of 0.44 the to positive 0.89 nC and·m− negative3 as the chargeinitial CCN densities with baseline sounding (without changing the water vapor condition), the peak value of the intuitivelyconcentr reflectation the increased intensity from of electrification 300 to 1500 c intensitym−3 and the in the negative thunderstorm. charge density In the cases changed with from baseline positive charge density increased significantly from 0.44 to 0.89 nC·m−3 as the initial CCN –0.86 to –1.47 nC·m−3 as the initial CCN concentration increased. Under conditions with ample soundingconcentr (withoutation increased changing from the 300 water to 1500 vapor cm condition),−3 and the the negative peak valuecharge of density the positive changed charge fromdensity amounts of water vapor, the electrical activity of−3 the thunderstorm was significantly enhanced with increased–0.86 to significantly–1.47 nC·m−3 fromas the 0.44initial to CCN 0.89 nC concentration·m as the increased initial.CCN Under concentration conditions with increased ample from increasing −concentrations3 of aerosols. As the vapor mixing ratio decreased, the peak values−3 of the 300amounts to 1500 of cm waterand vapor, the the negative electrical charge activity density of the changedthunderstorm from was –0.86 signifi to –1.47cantly nCenhanced·m as with the initial positive and negative charge densities also decreased with respect to the baseline sounding and were CCNincreasing concentration concentrations increased. of aerosol Unders. As conditions the vaporwith mixing ample ratio amountsdecreased, of the water peak vapor, values theof the electrical <0.1 nC·m−3 at vapor mixing ratios <94%. The response of the peak values of the positive and negative activitypositive of and the thunderstormnegative charge wasdensities significantly also decreased enhanced with respect with to increasing the baseline concentrations sounding and were of aerosols. charge densities to an increase in aerosol concentration became less significant with decreasing As< the0.1 nC vapor·m−3 at mixing vapor ratiomixing decreased, ratios <94%. the The peak response values of the of peak the positive values of and the positive negative and charge negative densities water vapor content. For low amounts of water vapor, increasing the aerosol concentration had charge densities to an increase in aerosol concentration became less significant−3 with decreasing also decreasedlittle impact with on the respect peak values to the of baseline the positive sounding and negative and were charge <0.1 densities nC·m . Hence,at vapor the mixingeffects of ratios water vapor content. For low amounts of water vapor, increasing the aerosol concentration had <94%.aerosol The responses on the intensity of the peak of electrification values of the were positive suppressed and negative at low amounts charge of densities water vapor, to an leading increase in little impact on the peak values of the positive and negative charge densities. Hence, the effects of aerosolto a concentration lower sensitivity became of the intensity less significant of electrifi withcation decreasing to change waters in aerosol vapor concentration. content. For low amounts aerosols on the intensity of electrification were suppressed at low amounts of water vapor, leading of waterto a lower vapor, sensitivity increasing of the intensity the aerosol of electrifi concentrationcation to change had littles in aerosol impact concentration. on the peak values of the

Atmosphere 2016, 7, 137 5 of 19

Atmosphere 2016, 7,137 5 of 19 positive and negative charge densities. Hence, the effects of aerosols on the intensity of electriﬁcation

wereAtmosphere suppressed 2016, 7,137 at low amounts of water vapor, leading to a lower sensitivity of the intensity5 of 19 of electriﬁcation to changes in aerosol concentration.

Figure 3. Peak values of (a) positive and (b) negative charge density (nC·m−3) at 18 min of simulation Figure 3. Peak values of (a) positive and (b) negative charge density (nC·m−3) at 18 min of simulation underFigure different 3. Peak vapor values conditions of (a) positive and andinitial (b) CCNnegative concentrations. charge density (nC·m−3) at 18 min of simulation under different vapor conditions and initial CCN concentrations. Thisunder study different considered vapor conditions the non and-inductive initial CCN electrification concentrations. process that results from rebounding collisionsThisThis of study studyice-graupel considered considered and the snowthe non-inductive non-graupel.-inductive The electrification electrification number of ice process process-phase that that particles result resultss directlyfrom from rebounding rebounding affects the intensitycollisionscollisions of of electrical of ice-graupel ice-graupel activity andand of snow-graupel. snow the -thunderstormgraupel. T Thehe number number[31–33] . of Figure of ice ice-phase-phase 4 shows particles particles the mean directly directly concentration affect affectss thethe intensity of electrical activity of the thunderstorm [31–33]. Figure 4 shows the mean concentration ofintensity ice crystals, of electrical snow particles, activity of and the thunderstormgraupel particles [31– at33 ].18 Figure min of4 showssimulation the mean under concentration different vapor of ice conditionsof ice crystals, and initialsnow particles,CCN concentrations. and graupel particles For the at 18 cases min withof simulation baseline under sounding, different the vapor mean crystals, snow particles, and graupel particles at 18 min of simulation under different vapor conditions concentrationsconditions andof ice initial crystals, CCN snow concentrations. particles, and For graupel the cases particle withs increased baseline by sounding, 12%, 40 % the, and mean 36%, and initial CCN concentrations. For the cases with baseline sounding, the mean concentrations of ice respectively,concentrations when of ice the crystals, initial CCNsnow particles, concentration and graupel was increased particles increased from 300 by to 12 1500%, 40 % cm, and−3. As36%, the crystals, snow particles, and graupel particles increased by 12%, 40%, and 36%, respectively, when aerosolrespectively, concentration when increased the initial, CCNmore concentrationice-phase particles was increased were produced from 300 in the to 1500thunderstorm cm−3. As the and the initial CCN concentration was increased from 300 to 1500 cm−3. As the aerosol concentration participateaerosol concentrationd in the collision increased separation, more processice-phase under particles condition were produceds with sufficient in the thunderstorm vapor, enhancing and increased, more ice-phase particles were produced in the thunderstorm and participated in the collision the participatethunderstormd in the electrification collision sep aration process. process Previous under studies condition [1–s3] with have sufficient suggested vapor that, enhancing as more separationthe thunderstorm process under electrification conditions process. with sufficient Previous vapor, studies enhancing [1–3] have the thunderstorm suggested that electrification as more ice-phase particles participate in the process of collision separation, the intensity of electrical process.ice-phase Previous particles studies participate [1–3] inhave the suggested process of that collision as more separation ice-phase, the particles intensity participate of electric inal the activity is enhanced. As the amount of water vapor decreases, the response of the concentrations processactivity of is collision enhanced separation,. As the amount the intensity of water of vapor electrical decreases activity, isthe enhanced. response Asof the the concentrations amount of water of ice crystals, snow particles, and graupel particles to changes in the aerosol concentration becomes vaporof ice decreases, crystals, snow the response particles, of and the graupel concentrations particles of to ice change crystals,s in snowthe aerosol particles, concentration and graupel becomes particles less significant, especially for the water vapor contents <91% of that for the baseline sounding. The toless changes significant, in the aerosolespecially concentration for the water becomes vapor con lesstent significant,s <91% of especially that for the for baseline the water sounding. vapor contents The effect of aerosol particles on the concentration of ice-phase particles is suppressed by decreasing the <91%effect of of that aerosol for the particles baseline on sounding. the concentration The effect of ice of aerosol-phase particles particles is on suppressed the concentration by decreasing of ice-phase the water vapor content, leading to a lower sensitivity of the intensity of electrical activity to a change particleswater vapor is suppressed content, leading by decreasing to a lower the sensitivity water vapor of the content, intensity leading of electric to aal lower activity sensitivity to a chang ofe the in the aerosol concentration. Because the electrification process is dominated by the collision intensityin the aerosol of electrical concentration activity. to Because a change the in electrification the aerosol concentration. process is dominated Because by the the electrification collision separation between ice-phase particles, the intensity of electrical activity, which is directly processseparation is dominated between by ice the-phase collision particles, separation the intensitybetween ice-phase of electric particles,al activity, the which intensity is ofdirectly electrical influenced by the number of ice-phase particles, becomes less sensitive to increasing aerosol activity,influenced which by is the directly number influenced of ice-phase by the particles, number ofbecomes ice-phase less particles, sensitive becomesto increasing less sensitive aerosol to concentrations at lower water vapor contents (Figure 3). increasingconcentration aerosols at concentrationslower water vapor at lower content waters (Figure vapor 3).contents (Figure3).

Figure 4. Cont.

Atmosphere 2016, 7, 137 6 of 19 Atmosphere 2016, 7,137 6 of 19

Atmosphere 2016, 7,137 6 of 19

FigFigureure 4 4.. MMeanean concentration concentration (in (in cm cm−3−) 3of) of(a) ( aice) ice crystals, crystals, (b ()b snow) snow particles, particles, and and (c ()c )graupel graupel particles particles atat 18 18 min min of of simulation simulation under under different different vapor vapor condit conditionsions and and initial initial CCN CCN concentrations. concentrations.

−3 TheFig effectsure 4. M ofean aerosols concentration on the (in intensitycm ) of (a) ofice lightning crystals, (b )at snow different particles, water and (vaporc) graupel contents particles and initial Theat 18 effects min of ofsimulation aerosols under on thedifferent intensity vapor ofcondit lightningions and at initial different CCN concentrations. water vapor contents and initial CCN concentrations is shown by the mean flash rate (Figure 5). In the cases with baseline sounding, CCN concentrations is shown by the mean flash rate (Figure5). In the cases with baseline sounding, the mean flash rate was higher and increased with increasing initial CCN concentration. The mean the meanThe flasheffects rate of aerosols was higher on the and intensity increased of lightning with increasing at different initial water CCN vapor concentration. contents and initial The mean flash rate was increased by about 16.7% as the initial CCN concentration changed from 300 to 1500 cm–3. flashCCN rate concentrations was increased is shown by about by 16.7%the mean as theflash initial rate ( CCNFigure concentration 5). In the cases changed with baseline from 300sounding, to 1500 cm–3. Thethe electrification mean flash rate processes was higher in thundersto and increasedrms withare caused increasing by reboundinginitial CCN concentration.collisions between The mean ice-phase The electrification processes in thunderstorms are caused by rebounding collisions between ice-phase particlesflash rate [34 was]. A increaseds the thunderstorm by about 16.7% develops as the initial, the CCNcharges concentration carried by changed ice-phase from particles 300 to 1500 accumulate cm–3. particles [34]. As the thunderstorm develops, the charges carried by ice-phase particles accumulate andThe this electrification will lead to processes an increase in thundersto in the electricrms are field caused in theby reboundingcloud. When collisions the electric between field ice exceeds-phase the and this will lead to an increase in the electric field in the cloud. When the electric field exceeds the dischargeparticles threshold, [34]. As the a thunderstorm flash is triggered develops and, thepart charges of the carriedcharge by is iceneutralized-phase particles; the electric accumulate field will discharge threshold, a flash is triggered and part of the charge is neutralized; the electric field will thenand be this below will thelead discharge to an increase threshold. in the electric The electric field in field the cloud.will accumulate When the electric again untilfield exceedsthe next the flash is then be below the discharge threshold. The electric field will accumulate again until the next flash triggereddischarge [23,24,28, threshold,31 ]a. Asflash the is watriggeredter vapor and contentpart of the decreased, charge is theneutralized mean flash; the electricrate decreased field will by at isthen triggered be below [23 ,the24, 28discharge,31]. As threshold. the water The vapor electric content field decreased,will accumulate the meanagain until flash the rate next decreased flash is by at least 86%. This illustrates the importance of water vapor content in the intensity of electrical activity. leasttriggered 86%. This[23,24,28, illustrates31]. As thethe importancewater vapor ofcontent water decreased, vapor content the mea in then flash intensity rate decreased of electrical by at activity. The increase in the flash rate with increasing aerosol concentrations weakened with a decrease in Theleast increase 86%. This in theillustrates flash rate the importance with increasing of water aerosol vapor concentrations content in the intensity weakened of electrical with a decrease activity. in the the water vapor content. At the lowest water vapor content (85% of the baseline sounding), the waterThe increase vapor content. in the flash At the rate lowest with waterincreasing vapor aerosol content concentrations (85% of the baselineweakened sounding), with a decrease the mean in flash mean flash rate remained almost constant with increasing aerosol concentrations. Previous ratethe remained water vapor almost conten constantt. At the with lowest increasing water aerosol vapor content concentrations. (85% of the Previous baseline studies sounding), [1–4] thehave also studies [1–4] have also indicated that an increase in the concentration of aerosol particles indicatedmean flash that rate an increaseremained in thealmost concentration constant with of aerosol increasing particles aerosol enhances concentration the intensitys. Previous of electrical enhancesstudies the[1–4] intensity have also of indicated electrical that activity. an increase However, in thisthe concentration does not apply of aerosol to all thunderstorms particles activity. However, this does not apply to all thunderstorms under different conditions, and there are underenhances different the conditions,intensity of and electrical there activity.are still However,uncertainties this in does the not effects apply of toaerosol all thunderstorms on thunderstorm still uncertainties in the effects of aerosol on thunderstorm microphysics and electrification processes. microphysicsunder different and conditions, electrification and there processes. are still At uncertainties low wate rin vapor the effects content of aerosols, the intensityon thunderstorm of electrical At low water vapor contents, the intensity of electrical activity is little changed with increasing aerosol activitymicrophysics is little and changed electrification with increasing processes. aerosolAt low wate concentrationr vapor contents, whichs, the showsintensity that of electrical the effects of concentrations, which shows that the effects of aerosol on thunderstorm electrical activity are heavily aerosolactivity on isthunderstorm little changed electrical with increasing activity are aerosol heavily concentration dependents on, which the amount shows ofthat water the vapor effects present of . dependentaerosol on thunderstorm on the amount electrical of water activity vapor are present. heavily dependent on the amount of water vapor present.

Figure 5. Mean flash rate (flash min–1)–1 for different vapor contents and initial CCN concentrations. FigFigureure 5 5.. MMeanean flash ﬂash rate rate (flash (ﬂash min min–1) for) for different different vapor vapor contents contents and and initial initial CCN CCN concentrations. concentrations.

Atmosphere 2016, 7, 137 7 of 19

Aerosol particles affect thunderstorm electrical activity by influencing the hydrometeors and related microphysical processes. Figure6 shows the variation in the mean number concentration of cloud droplets, raindrops, ice crystals, snow particles, and graupel particles with time for initial CCN concentration of 300 and 1500 cm−3. The concentration of cloud droplets in a thunderstorm significantlyAtmosphere increases 2016, 7,137 and the size of the droplets decreases with increasing concentrations7 of 1 of9 aerosols (Figure6a). In the simulations with the lowest water vapor content (85% of the baseline sounding), Aerosol particles affect thunderstorm electrical activity by influencing the hydrometeors and the increaserelated in the microphysical concentration processes. of cloud Figure droplets 6 shows causedthe variation by increasing in the mean initial number CCN concentration concentrations of was smaller thancloud in droplets, the simulations raindrops, icewith crystals, ample snow amount particles, of water and graupel vapor particles (100% with of the time baseline for initial sounding). Fewer dropletsCCN concentration were activated of 300 at and lower 1500 amounts cm−3. The of concentration water vapor, of whencloud droplets the concentration in a thunderstorm of aerosol was greater. Assignificantly the concentration increases and of aerosolthe size ofparticles the droplets increased, decreases the with size increasing of the cloud concentration dropletss of decreased aerosols (Figure 6a). In the simulations with the lowest water vapor content (85% of the baseline and the collisionsounding), and the coalescence increase in the processes concentration were of suppressed, cloud droplets resulting caused by in increasing fewer raindrops. initial CCN Figure 6b shows thatconcentration increasings was the smaller aerosol than concentration in the simulations had with a ample significant amount impact of water onvapor the (100% concentrations of the of cloud dropletsbaseline and sounding). raindrops Fewer afterdroplets 25 were min activated of simulation at lower with amounts an ampleof water amount vapor, when of water the vapor. The differenceconcentration in the of concentrationaerosol was greater of. As raindrops the concentration caused of by aerosol changes particles in increased the aerosol, the size concentration of the cloud droplets decreased and the collision and coalescence processes were suppressed, resulting became unremarkable at the lowest water vapor content (85% of the baseline sounding). Fan et al. [18] in fewer raindrops. Figure 6b shows that increasing the aerosol concentration had a significant showed thatimpact aerosol on the effectsconcentrations on precipitation of cloud droplets were and significant raindropsunder after 25 relatively min of simulation humid with conditions, an but became lessample significant amount of underwater vapor. conditions The difference with limited in the concentration water vapor. of raindrops The water caused vapor by contentchanges affected the concentrationin the aerosol of iceconcentration crystals, became which unremarkable decreased with at the decreasing lowest water water vapor vaporcontent content.(85% of the However, the influencebaseline of aerosol sounding). particles Fan et al. on [18] the showed concentration that aerosol of ice effects crystals on precipitation was less obviouswere significant and there was under relatively humid conditions, but became less significant under conditions with limited water little differencevapor. The in the water concentration vapor content of affected ice crystals the concentration at different of aerosol ice crystals, concentrations which decreased under with both ample water vapordecreasing (100% water of the vapor baseline content. sounding) However, the and influence limited of water aerosol vapor particles (85% on the of concentration the baseline of sounding) conditions.ice crystals For high was amountsless obvious of and water there vapor, was little the difference concentration in the concentration of snow particles of ice crystals increased at with increasingdifferent aerosol aerosol concentrations, concentrations but under changed both ample very water little vapor at the(100% lowest of the waterbaseline vapor sounding) content. and In this limited water vapor (85% of the baseline sounding) conditions. For high amounts of water vapor, study, the non-inductive electrification processes result from rebounding collisions of ice-graupel and the concentration of snow particles increased with increasing aerosol concentrations, but changed snow-graupel.very little Miller at theet al.lowest [35] and water Mansell vapor content. et al. [2 ] In suggested this study, that the graupel non-inductive particles electrification play a part in the electrificationprocesses processes result from and rebounding have a direct collisio impactns of ice on-graupel thunderstorm and snow- electricalgraupel. Miller activity. et al. Figure[35] and6 e shows that increasingMansell aerosol et al. [2] concentrationssuggested that graupel resulted particles in the play formation a part in the of moreelectrification graupel processes particles, and which is more significanthave a direct under impact the on conditions thunderstorm of ample electrical amounts activity. Figure of water 6e shows vapor. that In increasing the simulations aerosol with a concentrations resulted in the formation of more graupel particles, which is more significant under low waterthe vapor conditions content, of ample increasing amounts theof water aerosol vapor. concentration In the simulations had with little a low effect water on vapor theconcentration content, of graupel particles.increasing the aerosol concentration had little effect on the concentration of graupel particles.

Figure 6. Cont. Atmosphere 2016, 7, 137 8 of 19 Atmosphere 2016, 7,137 8 of 19

FigureFigure 6. Mean 6. M concentrationean concentration of (ofa) ( clouda) cloud droplet, droplet, (b ()b raindrops,) raindrops, ( c()c) ice ice crystals, crystals, ((dd)) snowsnow particles, and and (e) graupel particles over time for initial CCN of 300 and 1500 cm−3 under water vapor conditions (e) graupel particles over time for initial CCN of 300 and 1500 cm−3 under water vapor conditions of of 85% and 100% of baseline sounding. 85% and 100% of baseline sounding. These results are consistent with previous studies [36–42] in which increasing the initial CCN Theseconcentrations results arereduced consistent the droplet with size, previous enhanced studies condensation [36–42] growth, in which and increasing reduced the the effi initialciency CCN concentrationsof the collision reduced and coalescence the droplet of size,droplets. enhanced This led condensation to more small droplets growth, being and reducedtransported the above efficiency of thethe collision freezing and level coalescence and enhanced of droplets. the growth This of led ice to-phase more smallparticles droplets under being conditions transported of ample above the freezingamounts level of water and vapor. enhanced Under the conditions growth of of ice-phaselimited wat particleser vapor, under the increase conditions in the ofconcentration ample amounts of cloud droplets caused by aerosols became less significant and fewer small droplets were carried of water vapor. Under conditions of limited water vapor, the increase in the concentration of cloud above the freezing level, leading to fewer raindrops and ice-phase particles. Under conditions of dropletsample caused amounts by aerosols of water became vapor, less there significant was an obvious and fewer influence small of droplets aerosols were on carriedthe ice-phase above the freezingparticles level, and leading related to fewer microphysical raindrops processes, and ice-phase resulting particles. in a significant Under conditions change in of the ample electrical amounts of wateractivity vapor, of thunderstorms. there was an obviousIn addition, influence the weak of aerosolsice-phase on microphysical the ice-phase processes particles and and fewer related microphysicalice-phase particles processes, suppressed resulting the in influencea significant of aerosols change on in electricalthe electrical activity, activity leading of thunderstorms. to a lower In addition,effect of aerosols the weak on ice-phasethe intensity microphysical of electrical activity processes under limited and fewer water ice-phase vapor conditions particles (Figure suppressed 5). the influenceGraupel of aerosols particles on are electrical very important activity, leadingin electrification to a lower processes effect of [ aerosols35], and on aerosols the intensity have of electricaldifferent activity effects under on graupel limited particles water vaporfor different conditions amounts (Figure of water5). vapor. Figure 7 shows the mass Graupelconversion particles rates ofare graupel very importantparticles, including in electrification the accretion processes of snow [35 by], and cloud aerosols droplet haves to form different graupel, the accretion of ice crystals by rain to form graupel, the accretion of rain by graupel effects on graupel particles for different amounts of water vapor. Figure7 shows the mass conversion particles, and the accretion of snow by rain to form graupel particles. The conversion rate of the ratesaccretion of graupel of particles, snow by cloud including droplet thes to accretion form graupel of snow particles by cloud is significantly droplets to increased form graupel, with the accretionincreasing of ice aerosol crystals concentration by rain tos under form conditions graupel, with the accretionample amounts of rain of water by graupel vapor, but particles, is very and the accretionlow or almost of snow zero when by rain there to are form only graupel low amounts particles. of water The vapor. conversion Under conditions rate of the with accretion ample of snowamounts by cloud of dropletswater vapor, to formthe conver graupelsion particlesrate of the is accretion significantly of ice increasedcrystals by withrain to increasing form graupel aerosol concentrationsparticles increases under conditions with increasing with aerosolample amounts concentration of waters, mainly vapor, due but to isthe very larger low ice or crystalsalmost zero whenpresent there areunder only polluted low amounts condition ofs. waterThe effects vapor. of aerosol Under on conditions the conversion with amplerate under amounts conditions of water vapor,with the ample conversion amounts rate of ofwater the vapor accretion were of consiste ice crystalsnt with by previous rain to studies form graupel[43,44]. The particles conversion increases rate decreased at lower amounts of water vapor, leading to a lower sensitivity to increasing aerosol with increasing aerosol concentrations, mainly due to the larger ice crystals present under polluted concentrations. There were more graupel particles as the aerosol concentration increased with conditions. The effects of aerosol on the conversion rate under conditions with ample amounts of water ample amounts of water vapor, leading to an increase in the conversion rate for the accretion of rain vaporby were graupel. consistent The conversion with previous rate of studies accretion [43 ,44 of ]. rain The by conversion graupel was rate relatively decreased lower at lower for lower amounts of wateramounts vapor, of leadingwater vapor, to a lower resulting sensitivity in a lower to increasingresponse to aerosolthe aerosol concentrations. concentration There. When were there more graupelwere particles ample asamounts the aerosol of water concentration vapor, theincreased number of with snow ample particles amounts increase of waterd with vapor, increasing leading to an increaseaerosol in concentration the conversions; however, rate for thethe accretion concentration of rain of by raindropsgraupel. The decreased conversion under rate polluted of accretion of raincondition by graupels and was the relativelyconcentration lower of snow for lower was very amounts low, resulting of water in vapor, a lower resulting conversion in a rate lower for response the to theaccretion aerosol of concentration. snow by rain to When form graupel. there were There ample was no amounts obvious of effect water of the vapor, aerosol the concentration number of snow particleson the increased conversion with rate increasing of the accretion aerosol of snow concentrations; by rain at lower however, amounts the of water concentration vapor. of raindrops decreased under polluted conditions and the concentration of snow was very low, resulting in a lower conversion rate for the accretion of snow by rain to form graupel. There was no obvious effect of the aerosol concentration on the conversion rate of the accretion of snow by rain at lower amounts of water vapor. Atmosphere 2016, 7, 137 9 of 19 Atmosphere 2016, 7,137 9 of 19

(a) (b)

FigureFigure 7. Mass 7. Mass conversion conversion rates rates of of graupel graupel particlesparticles with time time for for initial initial CCN CCN of of300 300 and and 1500 1500 cm− cm3 −3 underunder water water vapor vapor conditions conditions of of 85% 85% andand 100%100% of baseline baseline sounding, sounding, including including (a) (thea) theaccretion accretion of of snowsnow by by cloud cloud droplets droplets to to form form graupel, graupel, (b) the the accretion accretion of of ice ice crystals crystals by byrain rain to form to form graupel, graupel, (c) the(c) accretionthe accretion of rain of rain by graupel,by graupel, and and (d) ( thed) the accretion accretion of of snow snow by by rain rain to to form form graupel. graupel.

Under conditions with ample amounts of water vapor, the conversion rates for the accretion of snowUnder by conditionscloud droplets with to ampleform graupel, amounts the of accretion watervapor, of ice crystals the conversion by rain to rates form for graupel, the accretion the of snowaccretion by cloudof rain droplets by graupel, to formand the graupel, accretion the of accretion snow by of rain ice to crystals form graupel by rain increased to form with graupel, theincreasing accretion ofaerosol rain by concentrations, graupel, and leading the accretion to more of snowgraupel. by rainWhen to formthe water graupel vapor increased content with increasingdecreased, aerosol the conversion concentrations, rates for leading graupel to were more suppressed graupel. When and the the effect water of vaporaerosol content concentration decreased, theon conversion the graupel rates particles for graupelwas also weresuppressed. suppressed Hence, andthe effect the effectof aerosol of aerosol concentration concentration on ice-phase on the graupelparticles particles depended was alsoon the suppressed. amounts of Hence, water vapor the effect present. of aerosol concentration on ice-phase particles depended on the amounts of water vapor present. 4.2. Response of Aerosol Effects on the Charge Structures to the Amount of Water Vapor Present 4.2. ResponseFigure of 8 Aerosol shows Effectsthe peak on thevalues Charge of positive Structures and to negative the Amount charge of Water density. Vapor In Presentthe simulations withFigure baseline8 shows sounding the peak (ample values amount of positive of water and negative vapor), chargethe change density. in aerosol In the simulationsconcentration with affected the charge structures. When the initial CCN concentration was 300 cm–3, the structure baseline sounding (ample amount of water vapor), the change in aerosol concentration affected the retained dipolar throughout the simulation. However, when the initial CCN concentration was 1500 charge structures. When the initial CCN concentration was 300 cm–3, the structure retained dipolar cm–3, the charge structure was dipolar in the initial stage of simulation and then a new negative –3 throughoutcharge region the simulation. appeared above However, the main when positive the initial charge CCN center concentration after 34 min of was simulation. 1500 cm When, the the charge structurewater wasvapor dipolar content in thedecreased initial stage to 94% of simulationof the baseline and then sounding, a new negativethe charge charge structure region of appeared the abovethunderstorm the main positivewas still chargedipolar centerat low afteraerosol 34 concentrations min of simulation. and there When was the also water a new vapor negative content decreasedcharge toregion 94% above of the the baseline main positive sounding, charge the center; charge however, structure the of negative the thunderstorm charge region was appeared still dipolar at lowlate aerosol and decreased. concentrations As the andwater there vapor was content alsoa de newcreased negative further charge to 85% region of the above baseline the sounding, main positive chargethe center;charging however, process of the thunderstorm negative charge appeared region late appearedr, and the latecharge and structure decreased. remained As the dipolar water as vapor contentthe initial decreased CCN further concentration to 85% ofincreased. the baseline Therefore, sounding, the thereduction charging in processwater vapor of thunderstorm content appearedsuppressed later, the and effects the charge of the aerosol structure concentration remained dipolaron the charge as the structure. initial CCN concentration increased.

Therefore, the reduction in water vapor content suppressed the effects of the aerosol concentration on the charge structure. Atmosphere 2016, 7, 137 10 of 19 Atmosphere 2016, 7,137 10 of 19

Figure 8. Cont.

AtmosphereAtmosphere 20162016,, 77,, 137 11 of 1919 Atmosphere 2016, 7,137 11 of 19

(a) (b) (a) (b) Figure 8. Time–height peak values of positive (solid line) and negative (red dashed line) charge FigureFigure 8.8. Time–heightTime–height peak peak values values of positive of positive (solid (solid line) line) and negative and negative (red dashed (red dashed line) charge line) density charge density (nC·m–3) for initial concentrations of CCN of 300 (a) and 1500 cm–3 (b) under water vapor (nCdensity·m–3 )(nC for·m initial–3) for concentrations initial concentrations of CCN ofof 300CCN (a )of and 300 1500 (a) cmand–3 1500(b) under cm–3 ( waterb) under vapor water conditions vapor conditions from 100% to 85% of the baseline sounding at 3% intervals (from top to bottom). fromconditions 100% tofrom 85% 100 of% the to baseline 85% of soundingthe baseline at 3% sounding intervals at (from 3% intervals top to bottom). (from top to bottom). The main charging mechanism considered in this study was the non-inductive electrification The main charging mechanism considered in this study was the non-inductive electrification resultingThe mainfrom charging rebounding mechanism collisions considered of ice-graupel in this studyand snow was- thegraupel. non-inductive The charge electriﬁcation structures resulting from rebounding collisions of ice-graupel and snow-graupel. The charge structures resultingdirectly depend from reboundinged on the polarity collisions of ofthe ice-graupel charge carried and snow-graupel.by the ice-phase The particles. charge structuresAs a result directly of the directly depended on the polarity of the charge carried by the ice-phase particles. As a result of the dependedlarge number on theof ice polarity crystals, of thethe chargecharging carried statesby of thethe ice-phaseice crystals particles. were analyzed. As a result Figure of the9 shows large large number of ice crystals, the charging states of the ice crystals were analyzed. Figure 9 shows numberthe peak of values ice crystals, of the positive the charging and negative states of thecharge ice crystalsdensity werecarried analyzed. by the ice Figure crystals.9 shows Under the ample peak the peak values of the positive and negative charge density carried by the ice crystals. Under ample valueswater vapor of the conditions positive and (baseline negative sounding), charge density the icecarried crystals by were the mainly ice crystals. distributed Under at ample a height water of water vapor conditions (baseline sounding), the ice crystals were mainly distributed at a height of vapor7–10 km conditions and were (baseline always sounding),positively charged the ice crystals when the were initial mainly CCN distributed concentration at a height was 300 of 7–10cm–3. kmAs 7–10 km and were always positively charged when the initial CCN concentration was–3 300 cm–3. As andthe initialwere always CCN concentrationpositively charged is increased when the to initial 1500 CCN cm−3 concentration, the ice crystals was were 300 cm distributed. As the initial more the initial CCN concentration is increased − to3 1500 cm−3, the ice crystals were distributed more CCNwidely concentration (at a height of is increased6–11.5 km) to, with 1500 the cm ice, thecrystals ice crystals becoming were positively distributed charge mored in widely the first (at 30 a widely (at a height of 6–11.5 km), with the ice crystals becoming positively charged in the first 30 heightmin of ofsimulation 6–11.5 km), and with some the of ice the crystals ice crystals becoming above positively a height charged of 8 km in becoming the ﬁrst 30 negatively min of simulation charged min of simulation and some of the ice crystals above a height of 8 km becoming negatively charged andlater. some This ofshows the ice the crystals charge above reversal a height of ice of crystals 8 km becoming in this region. negatively When charged the water later. vapor This content shows later. This shows the charge reversal of ice crystals in this region. When the water vapor content thewas charge 94% of reversal that in the of ice baseline crystals sounding, in this region. the ice When crystals the charg watered vapor positively content when was the 94% initial of that CCN in was 94% of that in the baseline sounding, the ice crystals charged positively when the initial CCN theconcentration baseline sounding, was 300 the cm− ice3, whereas crystals charged some of positivelythe ice crystals when the above initial a height CCN concentration of 8 km became was concentration−3 was 300 cm−3, whereas some of the ice crystals above a height of 8 km became 300negatively cm , whereascharged some with of increasing the ice crystals concentrations above a height of aerosol of 8 km. The became distribution negatively of charged the negative with negatively charged with increasing concentrations of aerosol. The distribution of the negative increasingcharge region concentrations decreased with of aerosol. decreasing The distributionamounts of ofwater the negativevapor. As charge the amount region decreasedof water vapor with charge region decreased with decreasing amounts of water vapor. As the amount of water vapor decreasingdecreased further amounts to of 85% water of that vapor. in Asthethe baseline amount sounding, of water the vapor distribution decreased of further the charge to 85% carried of that by in decreased further to 85% of that in the baseline sounding, the distribution of the charge carried by thethe baselineice crystals sounding, decreased the distribution and the ice of crystals the charge charge carriedd positively by the ice only. crystals Therefore, decreased the and negatively the ice the ice crystals decreased and the ice crystals charged positively only. Therefore, the negatively crystalscharged charged region that positively appeared only. above Therefore, the main the negatively positive char chargedge center region under that conditions appeared above of ample the charged region that appeared above the main positive charge center under conditions of ample mainamounts positive of water charge vapor center and under high conditions initial CCN of ampleconcentration amountss ofwa waters caused vapor by andthe highnegative initial charge CCN amounts of water vapor and high initial CCN concentrations was caused by the negative charge concentrationscarried by the wasice crystals.caused by Under the negative conditions charge with carried a low by water the ice vapor crystals. content, Under increasing conditions aerosol with a carried by the ice crystals. Under conditions with a low water vapor content, increasing aerosol lowconcentration water vapors did content, not lead increasing to a reversal aerosol of concentrationscharge carried didby notthe leadice particles to a reversal, resulting of charge in similar carried concentrations did not lead to a reversal of charge carried by the ice particles, resulting in similar bycharge the ice structures particles, between resulting clean in similar and polluted charge structures simulations between. clean and polluted simulations. charge structures between clean and polluted simulations.

Figure 9. Cont.

Atmosphere 2016, 7, 137 12 of 19 Atmosphere 2016, 7,137 12 of 19

(a) (b)

Figure 9. Time–height peak values of positive and negative charge density carried by ice crystals for initial concentrations of CCN of 300 (a) and 1500 cm–3 (b) under water vapor conditions from 100% to 85% of baseline sounding at 3% intervals (from top to bottom). Atmosphere 2016, 7,137 13 of 19

Figure 9. Time–height peak values of positive and negative charge density carried by ice crystals for initial concentrations of CCN of 300 (a) and 1500 cm–3 (b) under water vapor conditions from 100% to Atmosphere 2016, 7, 137 13 of 19 85% of baseline sounding at 3% intervals (from top to bottom).

The polarity of charges carried by ice particlesparticles after the reboundingrebounding collisioncollision depends on the difference between the RAR and the RARC [34]. If RAR > RARC, then the graupel particles carry difference between the RAR and the RARC [34]. If RAR > RARC, then the graupel particles carry positive charge andand thethe reboundingrebounding iceice crystalscrystals oror snowsnow particlesparticles carrycarry negativenegative charge.charge. When RAR < RARC, graupel particles carry negative charge, and the rebounding ice crystals or snow RAR < RARC, graupel particles carry negative charge, and the rebounding ice crystals or snow particles carrycarry positivepositive charge.charge. To analyzeanalyze the high-levelhigh-level negative charge carried by the ice crystals, the positive deviation of RAR and RARC was plotted in Figure 10. In the simulations with the the positive deviation of RAR and RARC was plotted in Figure 10. In the simulations with the baseline baseline sounding, the region of the positive deviation of RAR and RARC became broader with sounding, the region of the positive deviation of RAR and RARC became broader with increasing aerosolincreasing concentration, aerosol concentration, especially aboveespecially a height above of a 8 height km, which of 8 iskm, the which main regionis the main for rebounding region for collisionsrebounding between collisions ice between particles. ice This particles. led to aThis negative led to chargea negative carried charge by icecarried crystals by ice and crystals a positive and chargea positive carried charge by graupel carried particles by graupel in this particles region. When in this the region. water vapor When content the water decreased vapor to content 94% of decreased to 94% of the baseline sounding, the distribution of the positive deviation of RAR and the baseline sounding, the distribution of the positive deviation of RAR and RARC decreased. As the RARC decreased. As the water vapor content decreased further to 88% of the baseline sounding, the water vapor content decreased further to 88% of the baseline sounding, the distribution of the positive distribution of the positive deviation of RAR and RARC decreased further and was mainly deviation of RAR and RARC decreased further and was mainly distributed below a height of 7 km, distributed below a height of 7 km, where the collision of ice-phase particles was too low. There where the collision of ice-phase particles was too low. There was therefore no change in the charge was therefore no change in the charge polarity of ice particles with increasing aerosol concentration polarity of ice particles with increasing aerosol concentration when the amount of water vapor was low. when the amount of water vapor was low.

Figure 10. Cont.

Atmosphere 2016, 7, 137 14 of 19 Atmosphere 2016, 7,137 14 of 19

(a) (b)

–2 −1 FigFigureure 10.10. TimeTime–height–height positive deviation of RAR and RAR C (in(in g g·m·m–2·s·−s1) for) for initial initial concentrations concentrations of − CCNof CCN of of300 300 (a) ( aand) and 1500 1500 cm cm−3 (3b)( bunder) under the the water water vapor vapor conditions conditions from from 100 100%% to to 85% 85% of of baseline sounding at 3% intervals (from top to bottom).

In this model, RARC is a function of temperature and remains unchanged at the same In this model, RARC is a function of temperature and remains unchanged at the same temperature, whereastemperature, RAR whereas depends RAR on the depends liquid wateron the content liquid water (see Figure content 11 ).(see In Figure the simulations 11). In the with simulations baseline withsounding, baseline the distribution sounding, the of liquid distribution water was of liquid higher water with increasingwas higher aerosol with concentrations increasing aerosol and concentrationreached a heights and of 9.5 reach kmed after a height 30 min of of 9.5 simulation. km after In 30 polluted min of cases, simulation. the liquid In polluted water content cases, wasthe liquidgreater, water especially content at a was height greater of 6–8, km especially after 30 at min a ofheight simulation, of 6–8 because km after more 30 smallmin of droplets simulation, were becausetransported more to higher small levels. droplets As were the water transported vapor content to higher decreased levels to. As 94% the of the water baseline vapor sounding, content decreasedthe distribution to 94% of of liquid the baseline water contentsounding, became the distribution lower at a of height liquid of water 9 km, content although became an increasing lower at aconcentrations height of 9 ofkm, aerosol although increased an increasing the liquid waterconcentration content.s When of aerosol the water increase vapord contentthe liquid decreased water content.further to When 85% of the the water baseline vapor sounding, content the decreased distribution further of liquid to 85% water of wasthe baseline very close sounding, to that of the distribincreaseution in aerosol of liquid concentration water was very and close was mainlyto that foundof the increas below 8e in km. aerosol Under concentration conditions of and limited was watermainly vapor, found the below lower 8 sensitivitykm. Under of conditions the cloud of droplets limited to water the aerosol vapor, concentration the lower sensitivity resulted ofin the cloudliquid droplets water content to the above aerosol the concentration freezing level resulted remaining in constantthe liquid as water aerosol content concentration above the increased. freezing level remaining constant as aerosol concentration increased.

Atmosphere 2016, 7, 137 15 of 19 Atmosphere 2016, 7,137 15 of 19

Figure 11. Cont.

Atmosphere 2016, 7, 137 16 of 19 Atmosphere 2016, 7,137 16 of 19

(a) (b)

FigureFigure 11.11. TimeTime–height–height of liquid liquid water water content content for for initial initial concentrations ofof CCNCCN of 300 ( a) and 1500 cm −33 (b(b)) under under the the wate waterr vapor vapor conditions conditions from from 100 100%% to to 85% 85% of of the the baseline baseline sounding sounding at 3% intervals (from(from top top to to bottom). bottom).

Under conditions of ample amounts of water vapor, more cloud droplets were transportedtransported above the freezing level for polluted cases, leadingleading to a higher liquidliquid water content and a positive C deviation of of RAR RAR and and RAR RARC at highat high levels. level Ins. this In region, this region, a charging a charging reversal reversal was seen was for seen ice particles for ice duringparticles the during rebounding the rebounding collision process, collision resulting process, in the resulting graupel in becoming the graupel positively becoming charged, positively the ice crystals/snowcharged, the ice particles crystals/snow becoming parnegativelyticles becoming charged, andnegatively a small negativelycharged, chargedand a small region negative appearingly abovecharge thed region main positive appearing charge above center. the As main the water positive vapor charge content center. decreased, As the a lower water sensitivity vapor content of the clouddecreased, droplets a lower to an sensitivity increase in of the the aerosol cloud concentrationdroplets to an led increase to a smaller in the differenceaerosol concentration in the liquid led water to contenta smaller between difference the in clean the liquid and the water polluted content simulations between the with clean a lower and the distribution polluted simulations for liquid water. with C Therea lower was distribution no positive for deviation liquid water. of RAR There and RARwas Cnoin positive the region deviation with rebounding of RAR and collisions RAR in of the ice particles,region with resulting rebounding in no change collision ins the of charge ice particles structure, resulting with increasing in no change aerosol in concentrations. the charge structure Hence, thewith reduction increasing in aerosol the water concentration vapor contents. Hence, suppressed the reduction the effects in of the aerosols water vapor on the content charge structure.suppressed the effects of aerosols on the charge structure. 5. Summary and Conclusions 5. Summary and Conclusions Non-inductive electrification and a simple discharge process were coupled with Morrison two-momentNon-inductive bulk mixed-phase electrification microphysics and a simple scheme discharge in the WRFprocess model. were Thirty coupled susceptibility with Morrison tests weretwo-moment simulated bulk to exploremixed-phase the effects microphysics of aerosol scheme concentration in the WRF on the model. development Thirty susceptibility of electrification tests inwere thunderstorms simulated to under explore different the effects vapor of conditionsaerosol concentration by varying theon the initial development CCN concentration of electrification and the waterin thunderstorm vapor contents under below different the freezing vapor level. conditions by varying the initial CCN concentration and the waterOur resultsvapor content showed below that the the previously freezing level. reported [1–5] enhancement in the lightning activity of thunderstormsOur results as show the concentrationed that the previously of aerosols reported increases [1– is5] not enhancement universal. Underin the lightning conditions activity of ample of amountsthunderstorms of water as vapor, the concentration increasing the of aerosol aerosols concentration increases is resulted not universal. in the formation Under ofconditions more small of cloudample droplets,amounts reducing of water the vapor, collision increasing coalescence the aerosol efficiency concentration of the cloud resulted droplets andin the suppressing formation the of warmmore small rain process. cloud droplets, More small reducing droplets the were collision transported coalescence to above efficiency the freezing of the level cloud and thisdroplets enhanced and thesuppressing lightning theactivity warm resulting rain process from the. M participationore small droplets of more were ice particles transported in the to electrification above the freezing process. Aslevel the and water this vapor enhanced content the decreased, lightning theactivity sensitivity result ofing cloud fromdroplets, the participation rain droplets, of more and ice ice particles in the electrification process. As the water vapor content decreased, the sensitivity of cloud droplets,

Atmosphere 2016, 7, 137 17 of 19 to an increase in the aerosol concentration decreased, leading to smaller differences in the intensity of lighting activities between the clean and polluted simulations. Under conditions with ample amounts of water vapor, more cloud droplets were transported above the freezing level in polluted simulations, leading to the formation of more liquid water and the charge reversal of ice particles at high levels. This resulted in the formation of a negative charge region above the main positive charge center in the polluted simulations. Under conditions with low amounts of water vapor, the proﬁle of effective liquid water content in the polluted cases was close to that in the clean cases, resulting in a similar charge distribution between the clean and polluted simulations as a result of the low amounts of water vapor reducing the sensitivity of the liquid water content to the concentration of aerosol. A decreasing water vapor content below freezing level hinders the development of thunderstorms, especially in the ice-phase processes. The effects of aerosols on the electrical processes of thunderstorms depended on the liquid water content and the concentration of ice particles above the freezing level. Reducing the water vapor content below the freezing level weakened the ice-phase processes and then suppressed the effects of aerosols on the intensity of electriﬁcation and the charge structure. The effects of aerosol concentration on cloud microphysical processes are very complicated [45], resulting in different effects of aerosol concentrations on the electrical processes of thunderstorms at different stages of development with a strong dependency on the condition of water vapor below the freezing level.

Acknowledgments: Support from the National Basic Research and Development (973) Program of China (2014CB441403, 2014CB441401), the Scientific Research Foundation of CUIT (KYTZ201601), the Science and Technology Achievements Transformation Program of Sichuan Provincial Department of Education (16CZ0021), and the Key Program of Beijing Municipal Natural Science Foundation (8141002) is gratefully acknowledged. Author Contributions: Pengguo Zhao, Yan Yin, and Hui Xiao conceived and designed the experiments; Pengguo Zhao, Yunjun Zhou, and Jia Liu performed the experiments; Pengguo Zhao and Hui Xiao analyzed the simulation results and wrote the paper. Conflicts of Interest: The authors declare no conflict of interest.

References

1. Mitzeva, R.P.; Latham, J.; Petrova, S.A. comparative modeling study of the early electrical development of maritime and continental thunderstorms. Atmos. Res. 2006, 82, 26–36. [CrossRef] 2. Mansell, E.R.; Ziegler, C.L. Aerosol effects on simulated storm electrification and precipitation in a two-moment bulk microphysics model. J. Atmos. Sci. 2013, 70, 2032–2050. [CrossRef] 3. Zhao, P.; Yin, Y.; Xiao, H. The effects of aerosol on development of thunderstorm electrification: A numerical study. Atmos. Res. 2015, 153, 376–391. [CrossRef] 4. Bell, T.L.; Rosenfeld, D.; Kim, K.M.; Yoo, J.M.; Lee, M.I.; Hahnenberger, M. Midweek increase in U.S. summer rain and storm heights suggests air pollution invigorates rainstorms. J. Geophys. Res. 2008, 113.[CrossRef] 5. Wang, Y.; Wan, Q.; Meng, W.; Liao, F.; Tan, H.; Zhang, R. Long-term impacts of aerosols on precipitation and lightning over the Pearl River Delta megacity area in China. Atmos. Chem. Phys. 2011, 11, 12421–12436. [CrossRef] 6. Stallins, J.A.; Carpenter, J.; Bentley, M.L.; Ashley, W.S.; Mulholland, J.A. Weekend-weekday aerosols and geographic variability in cloud-to-ground lightning for the urban region of Atlanta, Georgia, USA. Reg. Environ. Chang. 2013, 1, 137–151. [CrossRef] 7. Middey, A.; Chaudhuri, S. The reciprocal relation between lightning and pollution and their impact over Kolkata, India. Environ. Sci. Pollut. Res. 2013, 20, 3133–3139. [CrossRef][PubMed] 8. Strikas, O.M.; Elsner, J.B. Enhanced cloud-to-ground lightning frequency in the vicinity of coal plants and highways in Northern Georgia, USA. Atmos. Sci. Lett. 2013, 14, 243–248. [CrossRef] 9. Kar, S.K.; Liou, Y.A. Enhancement of cloud-to-ground lightning activity over Taipei, Taiwan in relation to urbanization. Atmos. Res. 2014, 147, 111–120. [CrossRef] 10. Lyons, W.A.; Nelson, T.E.; Williams, E.R.; Cramer, J.A.; Turner, T.R. Enhanced positive cloud-to-ground lightning in thunderstorms ingesting smoke from fires. Science 1998, 282, 77–80. [CrossRef][PubMed] Atmosphere 2016, 7, 137 18 of 19

11. Murray, N.D.; Orville, R.E.; Huffines, G.R. Effect of pollution from Central American fires on cloud-to-ground lightning in May 1998. Geophys. Res. Lett. 2000, 27, 2249–2252. [CrossRef] 12. Fernandes, W.C.; Pinto, I.R.C.A.; Pinto, O., Jr.; Longo, K.M.; Freitas, S.R. New findings about the influence of smoke from fires on the cloud-to-ground lightning characteristics in the Amazon region. Geophys. Res. Lett. 2006, 33.[CrossRef] 13. Altaratz, O.; Koren, I.; Yair, Y.; Price, C. Lightning response to smoke from Amazonian fires. Geophys. Res. Lett. 2010, 37.[CrossRef] 14. Saleeby, S.M.; Cotton, W.R.; Lowenthal, D.; Borys, R.D.; Wetzel, M.A. Influence of cloud condensat ion nuclei on orographic snowfall. J. Appl. Meteorol. Climatol. 2009, 48, 903–922. [CrossRef] 15. Saleeby, S.M.; Cotton, W.R.; Fuller, J. The cumulative impact of cloud droplet nucleating aerosols on orographic snowfall in Colorado. J. Appl. Meteorol. Climatol. 2011, 50, 604–625. [CrossRef] 16. Saleeby, S.M.; Cotton, W.R.; Messina, J.; Beem, K. Aerosol impacts on the microphysical growth processes of orographic snowfall. J. Appl. Meteorol. Climatol. 2013, 52, 834–852. [CrossRef] 17. Tao, W.; Li, X.; Khain, A.; Matsui, T.; Lang, S.; Simpson, J. Role of atmospheric aerosol concentration on deep convective precipitation: Cloud-resolving model simulation. J. Geophys. Res. 2007, 112.[CrossRef] 18. Fan, J.; Zhang, R.; Li, G.; Tao, W. Effects of aerosols and relative humidity on cumulus clouds. J. Geophys. Res. 2007, 112.[CrossRef] 19. Stevens, B.; Feingold, G. Untangling aerosol effects on clouds and precipitation in a buffered system. Nature 2009, 461, 607–613. [CrossRef][PubMed] 20. Carrió, G.G.; Cotton, W.R. On the buffering of CCN impacts on wintertime orographic clouds: An idealized examination. Atmos. Res. 2014, 137, 136–144. [CrossRef] 21. Morrison, H.; Curry, J.A.; Khvorostyanov, V.I. A new double-moment microphysics scheme for application in cloud and climate models. Part 1: Description. J. Atmos. Sci. 2005, 62, 1665–1677. [CrossRef] 22. Morrison, H.; Gettelman, A. A new two-moment bulk strati-form cloud microphysics scheme in the Community Atmosphere Model, version 3 (CAM3), Part I: Description and numerical tests. J. Climate 2008, 21, 3642–3659. [CrossRef] 23. Mansell, E.R.; MacGorman, D.R.; Ziegler, C.L.; Straka, J.M. Charge structure and lightning sensitivity in a simulated multicell thunderstorm. J. Geophys. Res. 2005, 110, 1545–1555. [CrossRef] 24. Marshall, T.C.; MacCarthy, M.P.; Rust, W.D. Electric field magnitudes and lightning initiation in thunderstorms. J. Geophys. Res. 1995, 100, 7097–7103. [CrossRef] 25. Weisman, M.L.; Klemp, J.B. The dependence of numerically simulated convective storms on vertical wind shear and buoyancy. Mon. Weather Rev. 1982, 110, 504–520. [CrossRef] 26. Chu, C.M.; Lin, Y.L. Effects of orography on the generation and propagation of mesoscale convective systems in a two-dimensional conditionally unstable flow. J. Atmos. Sci. 2000, 57, 3817–3837. [CrossRef] 27. Gilmore, M.S.; Straka, J.M.; Rasmussen, E.N. Precipitation and evolution sensitivity in simulated deep convective storms: Comparisons between liquid-only and simple ice and liquid phase microphysics. Mon. Weather Rea. 2004, 132, 1897–1916. [CrossRef] 28. Mansell, E.R.; MacGorman, D.R.; Ziegler, C.L.; Straka, J.M. Simulated three-dimensional branched lightning in a numerical thunderstorm model. J. Geophys. Res. 2002, 107.[CrossRef] 29. Morrison, H.; Thompson, G.; Tatarskii, V. Impact of cloud microphysics on the development of trailing stratiform precipitation in a simulated squall line: Comparison of one- and two-moment schemes. Mon. Wea. Rea. 2009, 137, 991–1007. [CrossRef] 30. Seifert, A.; Beheng, K.D. A two-moment cloud microphysics parameterization for mixed- phase clouds. Part 2: Maritime vs. continental deep convective storms. Meteor. Atmos. Phys. 2006, 92, 67–82. [CrossRef] 31. MacGorman, D.R.; Cohen, A.; Mansell, T.; Ziegler, C. Flash rate, electrical, microphysical, and dynamical relationships across a simulated storm spectrum. In Proceedings of the 2008 AGU Fall Meeting, San Francisco, CA, USA, 15–19 December 2008. 32. Wiens, K.C. Summary of Colorado State University’s LMA-Related Observations from STEPS. In Proceedings of the 2008 AGU Fall Meeting, San Francisco, CA, USA, 15–19 December 2008. 33. Yair, Y.; Lynn, B.; Price, C.; Kotroni, V.; Lagouvardos, K.; Morin, E.; Mugnai, A.; Llasat, M.C. Predicting lightning density in Mediterranean storms based on the WRF model dynamic and microphysical fields. J. Geophys. Res. 2010, 115, D04025. [CrossRef] Atmosphere 2016, 7, 137 19 of 19

34. Saunders, C.P.R.; Peck, S.L. Laboratory studies of the inﬂuence of the rime accretion rate on charge transfer during crystal/graupel collisions. J. Geophys. Res. 1998, 103, 13949–13956. [CrossRef] 35. Miller, K.; Gadian, A.; Saunders, C.P.R.; Latham, J.; Christian, H. Modelling and observations of thundercloud electriﬁcation and lightning. Atmos. Res. 2001, 58, 89–115. [CrossRef] 36. Andreae, M.O.; Rosenfeld, D.; Artaxo, P.; Costa, A.A.; Frank, G.P.; Longo, K.M.; Silva-Dias, M.A.F. Smoking rain clouds over the Amazon. Science 2004, 303, 1337–1342. [CrossRef][PubMed] 37. Wang, C. A modeling study of the response of tropical deep convection to the increase of cloud condensation nuclei concentration: 1. Dynamics and microphysics. J. Geophys. Res. 2005, 110.[CrossRef] 38. Khain, A.; Rosenfeld, D.; Pokrovsky, J. Aerosol impact on the dynamics and microphysics of deep convective clouds. Q. J. R. Meteorol. Soc. 2005, 131, 2639–2663. [CrossRef] 39. Koren, I.; Kaufman, Y.J.; Rosenfeld, D.; Remer, L.A. Aerosol invigoration and restructuring of Atlantic convective clouds. Geophys. Res. Lett. 2005, 32.[CrossRef] 40. Van Den Heever, S.C.; Carrio, G.G.; Cotton, W.R.; DeMott, P.J.; Prenni, A.J. Impacts of nucleating aerosol on Florida storms. Part I: Mesoscale simulations. J. Atmos. Sci. 2006, 63, 1752–1775. [CrossRef] 41. Lin, J.C.; Matsui, T.; Pielke, S.R.A.; Kummerow, C. Effects of biomass-burning-derived aerosols on precipitation and clouds in the Amazon Basin: A satellite-based empirical study. J. Geophys. Res. 2006, 111. [CrossRef] 42. Chen, J.; Yin, Y.; Xiao, H.; Chen, Q. A numerical investigation of the impacts of anthropogenic sulfate aerosol on regional climate in East Asia. Asia-Pac. J. Atmos. Sci. 2014, 50, 391–403. [CrossRef] 43. Xiao, H.; Yin, Y.; Chen, Q.; Zhao, P. Impact of aerosol and freezing level on orographic clouds: A sensitivity study. Atmos. Res. 2016, 176, 19–25. [CrossRef] 44. Xiao, H.; Yin, Y.; Jin, L.; Chen, Q.; Chen, J. Simulation of the effects of aerosol on mixed-phase orographic clouds using the WRF model with a detailed bin microphysics scheme. J. Geophys. Res. 2015, 120, 8345–8358. [CrossRef] 45. Tao, W.; Chen, J.; Li, Z.; Wang, C.; Zhang, C. Impact of aerosols on convective clouds and precipitation. Rev. Geophys. 2012, 50.[CrossRef]

© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).