Observational evidence of aerosol enhancement of lightning activity and convective invigoration

Transcription

1 GEOPHYSICAL RESEARCH LETTERS, VOL. 38,, doi: /2010gl046052, 2011 Observational evidence of aerosol enhancement of lightning activity and convective invigoration Tianle Yuan, 1,2 Lorraine A. Remer, 2 Kenneth E. Pickering, 2 and Hongbin Yu 2,3 Received 2 November 2010; revised 3 January 2011; accepted 11 January 2011; published 16 February [1] Lightning activity over the West Pacific Ocean east of the Philippines is usually much less frequent than over the nearby maritime continents. However, in 2005 the Lightning Imaging Sensor (LIS) aboard the TRMM satellite observed anomalously high lightning activity in that area. In the same year the Moderate resolution Imaging Spectroradiometer (MODIS) measured anomalously high aerosol loading. The high aerosol loading was traced to volcanic activity, and not to any factor linked to meteorology, disentangling the usual convolution between aerosols and meteorology. We show that in general lightning activity is tightly correlated with aerosol loadings at both inter annual and biweekly time scales. We estimate that a 60% increase in aerosol loading leads to more than 150% increase in lightning flashes. Aerosols increase lightning activity through modification of cloud microphysics. Cloud ice particle sizes are reduced and cloud glaciation is delayed to colder temperature when aerosol loading is increased. TRMM precipitation radar measurements indicate that anomalously high aerosol loading is associated with enhanced cloud mixed phase activity and invigorated convection over the maritime ocean. These observed associations between aerosols, cloud microphysics, morphology and lightning activity are not related to meteorological variables or ENSO events. The results have important implications for understanding the variability of lightning and resulting aerosol chemistry interactions. Citation: Yuan, T., L. A. Remer, K. E. Pickering, and H. Yu (2011), Observational evidence of aerosol enhancement of lightning activity and convective invigoration, Geophys. Res. Lett., 38,, doi: /2010gl Introduction [2] There exists a strong land ocean contrast in lightning activity [Zipser, 1994; Williams and Stanfill, 2002]. This contrast can be largely traced to the difference in convective vigor between maritime and continental clouds [Williams and Stanfill, 2002]. In addition to the prominent thermodynamic and dynamic contrast between continental and oceanic atmosphere land ocean contrast in aerosols is also hypothesized to be an important factor. Indeed this hypothesis is supported by modeling and observational evidence that suggests aerosols play an important role in regulating convective strength [Andreae et al., 2004; Koren et al., 2005; 1 Joint Center for Environmental Technology UMBC, Baltimore, Maryland, USA. 2 Laboratory for Atmospheres, NASA Goddard Space Flight Center, Greenbelt, Maryland, USA. 3 Earth System Science Interdisciplinary Center, University of Maryland, College Park, Maryland, USA. Copyright 2011 by the American Geophysical Union /11/2010GL Wang, 2005; Zhang et al., 2007]. However, the role by aerosols is still debated largely due to the convolution of meteorology and aerosols [Lyons et al., 1998; Williams and Stanfill, 2002; Williams et al., 2002; Williams, 2005]. [3] In this study, we present observational support that aerosols can enhance lightning activity through modifying cloud microphysical properties and invigorating convection over the tropical ocean by using a variety of satellite measurements. We are confident in the separation of aerosol effect from that of meteorology because we trace the source of the enhanced aerosol loading to volcanic activity [ hvo.wr.usgs.gov/cnmi/] that is entirely independent of the meteorological factors. 2. Method and Data [4] The region of interest is the western Pacific east of the Philippines (5 N 20 N, 125 E 150 E). The region is typical of the tropical ocean with active convection, scarce lightning and pristine aerosol conditions. Total lightning flash count data from the Lightning Imaging Sensor (LIS) instrument aboard the Tropical Rainfall Measuring Mission (TRMM) is the primary data source of lightning activity [Boccippio et al., 2002; Christian et al., 2003] and it has a detection efficiency of 75 90%. The Precipitation Radar (PR) also aboard TRMM provides measurements of 3 D cloud precipitation fields. We use only convective cloud echoes whose top echo heights exceed 5 km. Aerosol loading is characterized by aerosol optical depth (AOD) retrieved by the Moderate resolution Imaging Spectroradiometer (MODIS) aboard the Terra and Aqua satellites. We use daily and monthly Level 3 Collection 5 aerosol optical depth at 550 nm over ocean [Remer et al., 2005]. The uncertainty of MODIS AOD follows Dt = ±0.03 ± 0.05t, where t is AOD. Clouds are also characterized by cloud top particle effective radius and cloud top temperature retrieved by MODIS aboard the Terra and Aqua satellites. We use daily Level 2 quality controlled retrievals [Platnick et al., 2003]. SO 2 (precursor for sulfate aerosol) column amounts [Krotkov et al., 2006] come from the Ozone Monitoring Instrument (OMI) onboard the Aura satellite. [5] We use vertical radar reflectivity profile (VRRP) constructed based on an ensemble of TRMM PR measurements [Zipser and Lutz, 1994] to qualitatively measure the mixed phase activity inside storms, which is important in lightning generation [Takahashi, 1978; Williams et al., 1991; Zipser and Lutz, 1994]. Similarly, cloud top temperature versus particle size profiles (T PS) constructed with an ensemble data of MODIS cloud product [Yuan and Li, 2010; Yuan et al., 2010] provide information on the cloud glaciation process and ice particle size, both of which are important 1of5

2 Figure 1. (a) MODIS AOD time series for our study area from June, July and August (Terra, solid line with triangles) and (Aqua, dashed line with squares); the horizontal long dashed line is the Terra climatological mean AOD and two dotted lines are +/ standard deviation of mean. (b) Time series of total lightning flash counts, solid line with triangles, for June, July, and August of 1998 through 2007 as measured by the LIS sensor on TRMM. Dashed horizontal line is the climatological mean and the two dotted lines are for +/ one standard deviation. (c) Total seasonal flash counts from TRMM LIS plotted against total seasonal areal mean AOD from MODIS for the overlapping period of (d) Same as in Figure 1c, but each triangle in this plot represents a bi weekly average of AOD and total of flash counts. for lightning generation [Takahashi, 1978; Williams et al., 1991; Zipser and Lutz, 1994; Sherwood et al., 2006]. 3. Results [6] Figure 1a shows the interannual variability of seasonal (June, July and August, JJA) and areal mean AOD calculated from the MODIS Level 3 daily data. We note an anomaly in areal mean AOD over our study area during JJA AOD increased by two standard deviations from the long term mean of around 0.1, or 60% increase above the long term mean. Correspondingly, Figure 1b shows that the total flash count from LIS data also increased two standard deviations above its climatological mean value during JJA of 2005, or an increase of 150%. Over the period of sensor overlap ( ) the JJA AOD and JJA flash counts are well correlated (r2 = 0.9, p < 0.001) as shown in Figure 1c. The correlation remains very robust (r2 = 0.84, p < 0.001) even if 2005 is removed from the calculation. These data suggest a link between aerosol concentration and lightning activity of deep convective clouds over this area. As shown in Figure 1d, on a biweekly time scale, the correlation between flash counts and AOD is still statistically significant at 99.9% confidence level. This relationship is thus neither a function of interannual variability of large scale synoptic conditions nor an accidental result of a few extremely active periods in the anomalous year of [7] Figure 2a shows the interannual variability of the retrieved cloud microphysics during the same period as in Figure 1 and the change in glaciation height, which is key to determining the extent of the mixed phase and thus lightning activity. Cloud particle size versus temperature (T PS) profiles constructed from a large ensemble of MODIS cloud retrievals show signs of outstanding features in The 2005 profile has the smallest maximum cloud particle size, suggesting an important microphysical modification of clouds has taken place that year. This decrease in cloud particle size is consistent with the observed increase in lightning activity [Sherwood et al., 2006]. In addition, glaciation temperature (the peak of crystal size in the T PS profiles) in 2005 was the coldest among all the years and is colder than the warmest year by 8 C. Such temperature difference translates into a 1 2 km increase in height of the glaciation level. Lower glaciation temperature (higher glaciation height) is typical of continental clouds and suggests stronger convective strength and active mixed phase process [Takahashi, 1978]. [8] The inference is corroborated with the interannual variability of the convective strength as measured by TRMM radar, shown in Figure 2b. In this plot, the radar reflectivity of the largest 1% of measured radar echoes is 2of5

3 Figure 2. (a) Temperature versus cloud particle size profiles for June, July and August from Aqua MODIS. Each profile is calculated from an ensemble of cloud top properties observed over the region. The 2005 profile is shown in solid line. The horizontal lines show the maximum cloud particle size, usually indicative of the top of the mixed phase layer, where full glaciation occurs. The 2005 glaciation level is shown by the solid horizontal line, and is at a significantly colder temperature. The extended mixed phase layer in 2005 is conducive for lightning activity. (b) Vertical radar reflectivity profiles from TRMM (June, July and August ): the most reflective 1% of radar echoes at each height are used to construct these profiles; the profile for 2005 is highlighted by the solid line. (c) Mixed phase strength against MODIS AOD: The difference in reflectivity between the levels 5 km and 10 km of the profiles shown in Figure 2c plotted as a function of the areal seasonal mean AOD retrieved from MODIS for the period between 2000 and The smaller the difference between these two levels, the more intense the convection. (d) SO 2 map for June 14, 2005 (in DU). The volcano is marked by the red dot and cross at around 16N 145E. A clear SO 2 plume is visible from the volcano stretching westward. A patch of high SO 2 from an earlier venting is also present to the east of Philippines. (e) Combined Terra and Aqua MODIS AOD map for June 14, When both instruments have valid retrievals we average the values. Otherwise, we complement one to the other. We note that some old SO 2 plume has been converted to sulfate as seen just to the east of Philippines. The new SO 2 plume (farther east) has not fully been converted given the low aerosol optical depth and high SO 2 concentration. averaged. It is these extreme events that are responsible for most lightning flashes [Zipser and Lutz, 1994]. The VRRP of 2005 stands out as substantially different and suggests stronger convection [Zipser and Lutz, 1994] than all other years, including recent moderately strong El Nino and La Nina events. We look at the profile between 5 km and 10 km, altitude range covering the depth of mixed phase for typical tropical deep convection. A prominent feature of lightning producing, continental VRRPs is their low lapse rates of radar reflectivity within the mixed phase depth, roughly characterized here by DZ 5 10 =Z 5 Z 10 ignoring the nonlinearity of VRRP between 5 km and 10 km. Lower DZ 5 10 implies more supercooled water content with large graupel transported across the freezing level and thus stronger riming activity. We note that 2005 has the lowest DZ 5 10 (Figure 2c), which is favorable for producing lightning and consistent with the direct flash count measurements. The tight negative relationship (p < 0.01) between DZ 5 10 and AOD (Figure 2c) closes the circle between lightning, convection, microphysics and aerosols. [9] Based on the above analysis and results from previous studies, the conceptual model of associations among aerosols, convection and lightning emerges as follows. Enhanced aerosol loading modifies the cloud microphysics by creating more, but smaller droplets. These numerous and small droplets suppress warm rain [Rosenfeld, 2000], lead to enhanced release of latent heat at higher altitudes, invigorate convection (Figures 2b and 2c) [Khain et al., 2004] and elevate glaciation levels (Figures 2a and 2c). The resulting changes in the mixed phase layer of the cloud enhance lightning activity (Figure 1). The aerosol effect on the cloud mixed phase activity between 5 and 10 km is central to this hypothesis (Figure 2c), and we see the strong negative correlation between DZ 5 10 and AOD or the enhanced mixed phase activity by aerosol in Figure 2c. This direct link between aerosol concentration and cloud mixed phase activity strongly supports the hypothesis 3of5

4 Table 1. Equivalent Potential Temperature and Cloud Base Height For Our Study Area From 1998 to 2007 Using Data From NCEP Reanalysis Data a Year e (K) CBH(m) a The values are calculated using data at 925 mb. As with ENSO indices 2005 is not anomalous in terms of these meteorological variables. pointing to enhanced aerosol loading as the cause of the increased lightning for our study area. Recent modeling studies with detailed electrification physics and/or cloud microphysics [Khain et al., 2008; Mansell et al., 2010] seem to support this conceptual model. [10] There remains the possibility that AOD is used here as a tracer of more traditional meteorological parameters that are the actual cause of the invigorated convection and enhanced lightning activity. For example, boundary layer humidity and temperature are instrumental in determining the efficiency of lightning production for convective clouds as shown by previous studies [Williams and Renno, 1993; Williams and Satori, 2004; Williams, 2005]. We examined several key environmental parameters from reanalysis data that are associated with strength of convection. As shown in Table 1, none of these parameters exhibits an anomaly in 2005 that would cause the observed large changes in the cloud microphysics and lightning activity. Only the aerosol is sufficiently different as discussed later. We also examined satellite observed temperature and relative humidity sounding data and found that the thermodynamics of the air mass is close to average for El Nino and Southern Oscillation (ENSO) may affect lightning distribution [Yoshida et al., 2007]. However, all ENSO indices, taking into consideration of variety of conditions such as sea surface temperature and surface pressure, point to a neutral summer in Furthermore, we had moderately strong to strong El Nino or La Nina events in 1998, 2000, 2002, 2004 and 2006 and neither of them stands out like 2005, a neutral year. No significant statistical relationship between ENSO indices and lightning activity over our study region is found. [11] More importantly, we find that the anomalous source of aerosols in 2005 is sulfate produced from SO 2 emitted from the Anatahan volcano [ on the Northern Mariana Islands (Figures 2d and 2e). An example is illustrated in Figures 2d and 2e where a plume of positive AOD anomaly originates from the volcano. These aerosols are not tracers for any anomalous air mass from over land. There are no anomalous cross equatorial flows that bring high biomass burning aerosols. Carbon monoxide is a tracer for both anthropogenic pollution and biomass burning. Its concentration is at a normal level in 2005 as well. These facts provide further confidence that observed link between aerosol concentration and lightning activity are the result of the physical processes outlined in previous paragraphs. 4. Discussion [12] Our results suggest that the quiescent maritime tropics have the potential to become more active in producing lightning when introduced with more aerosols. Here we explore a natural laboratory where the aerosol input originates from volcanic emissions and is decoupled from the meteorology. However, the observed responses in the clouds and lightning production are expected from aerosol particles introduced by anthropogenic sources, as well. Given the significant changes in source strength of tropical continental anthropogenic aerosols over the past decades from biomass burning and increased economic development upwind of major oceanic convection areas, the ramifications of aerosol impacts on various aspects of cloud properties such as lightning activity, convective strength and the vertical structure of latent heat release needs to be addressed. [13] However, sensitivity of convective cloud properties to aerosol concentration is a non linear function of background aerosol loading [Wang, 2005; Fan et al., 2009; Khain, 2009]. The background aerosol loading of our study area is among the lowest over the globe, which may make this area particularly sensitive to the addition of aerosols. Over much of the tropical maritime ocean deep tropical maritime convection occurs regularly, but lightning activity is much scarcer over ocean than those land based convection over land [Zipser, 1994; Williams and Stanfill, 2002]. Traditionally, this paradox is explained by land ocean contrasts in atmospheric thermodynamics and dynamics [Williams and Stanfill, 2002; Williams, 2005]. Another hypothesis hinges upon the land ocean contrast in aerosol loading and aerosol effect on cloud microphysics [Williams and Stanfill, 2002; Williams et al., 2002; Williams, 2005]. The results of this study tend to support the second hypothesis that the lack of aerosols over pristine ocean contributes to the absence of lightning activity relative to that observed over land. [14] Here we document a chain of events that links increased aerosol with changes to cloud microphysics and structure, leading to enhanced lightning activity. We expect that the enhanced lightning activity seen here will lead to observable changes in atmospheric chemistry, including tropospheric ozone production resulting from lightning NO x generation. [15] Acknowledgments. We thank J. V. Martins, I. Koren, C. Wang, M. Chin, Dale Allen, and R. Albrecht for helpful comments and suggestions. This research is supported by NASA s Interdisciplinary Research program. References Andreae, M. O., D. Rosenfeld, P. Artaxo, A. A. Costa, G. P. Frank, K. M. Longo, and M. A. F. Silva Dias (2004), Smoking rain clouds over the Amazon, Science, 303(5662), , doi: /science Boccippio, D. J., W. J. Koshak, and R. J. Blakeslee (2002), Performance assessment of the optical transient detector and lightning Imaging sensor. Part I: Predicted diurnal variability, J. Atmos. Oceanic Technol., 19(9), , doi: / (2002)019<1318:paotot>2.0. CO;2. 4of5

5 Christian, H. J., et al. (2003), Global frequency and distribution of lightning as observed from space by the Optical Transient Detector, J. Geophys. Res., 108(D1), 4005, doi: /2002jd Fan, J., T. Yuan, J. M. Comstock, S. Ghan, A. Khain, L. R. Leung, Z. Li, V. J. Martins, and M. Ovchinnikov (2009), Dominant role by vertical wind shear in regulating aerosol effects on deep convective clouds, J. Geophys. Res., 114, D22206, doi: /2009jd Khain, A. P. (2009), Notes on state of the art investigations of aerosol effects on precipitation: A critical review, Environ. Res. Lett., 4(1), , doi: / /4/1/ Khain, A., A. Pokrovsky, M. Pinsky, A. Seifert, and V. Phillips (2004), Simulation of effects of atmospheric aerosols on deep turbulent convective clouds using a spectral microphysics mixed phase cumulus cloud model. Part I: Model description and possible applications, J. Atmos. Sci., 61(24), , doi: /jas Khain, A., N. Cohen, B. Lynn, and A. Pokrovsky (2008), Possible aerosol effects on lightning activity and structure of hurricanes, J. Atmos. Sci., 65(12), , doi: /2008jas Koren,I.,Y.J.Kaufman,D.Rosenfeld,L.A.Remer,andY.Rudich (2005), Aerosol invigoration and restructuring of Atlantic convective clouds, Geophys. Res. Lett., 32, L14828, doi: /2005gl Krotkov, N. A., S. A. Carn, A. J. Krueger, P. K. Bhartia, and K. Yang (2006), Band residual difference algorithm for retrieval of SO 2 from the aura Ozone Monitoring Instrument (OMI), IEEE Trans. Geosci. Remote Sens., 44(5), , doi: /tgrs Lyons, W. A., T. E. Nelson, E. R. Williams, J. A. Cramer, and T. R. Turner (1998), Enhanced positive cloud to ground lightning in thunderstorms ingesting smoke from fires, Science, 282(5386), 77 80, doi: / science Mansell, E. R., C. L. Ziegler, and E. C. Bruning (2010), Simulated electrification of a small thunderstorm with two moment bulk microphysics, J. Atmos. Sci., 67(1), , doi: /2009jas Platnick, S., M. D. King, S. A. Ackerman, W. P. Menzel, B. A. Baum, J. C. Riedi, and R. A. Frey (2003), The MODIS cloud products: Algorithms and examples from Terra, IEEE Trans. Geosci. Remote Sens., 41(2), , doi: /tgrs Remer, L. A., et al. (2005), The MODIS aerosol algorithm, products, and validation, J. Atmos. Sci., 62(4), , doi: /jas Rosenfeld, D. (2000), Suppression of rain and snow by urban and industrial air pollution, Science, 287(5459), , doi: /science Sherwood, S. C., V. T. J. Phillips, and J. S. Wettlaufer (2006), Small ice crystals and the climatology of lightning, Geophys. Res. Lett., 33, L05804, doi: /2005gl Takahashi, T. (1978), Rimming electrification as a charge generation mechanism in thunderstorms, J. Atmos. Sci., 35(8), , doi: / (1978)035<1536:REAACG>2.0.CO;2. Wang, C. (2005), 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., 110, D21211, doi: /2004jd Williams, E. R. (2005), Lightning and climate: A review, Atmos. Res., 76(1 4), , doi: /j.atmosres Williams, E., and N. Renno (1993), An analysis of the conditional instability of the tropical atmosphere, Mon. Weather Rev., 121(1), 21 36, doi: / (1993)121<0021:aaotci>2.0.co;2. Williams, E. R., and G. Satori (2004), Lightning, thermodynamic and hydrological comparison of the two tropical continental chimneys, J. Atmos. Sol. Terr. Phys., 66(13 14), , doi: /j.jastp Williams, E., and S. Stanfill (2002), The physical origin of the land ocean contrast in lightning activity, C. R. Phys., 3(10), , doi: /s (02)01407-x. Williams, E. R., R. Zhang, and J. Rydock (1991), Mixed phased microphysics and cloud electrification, J. Atmos. Sci., 48(19), , doi: / (1991)048<2195:mpmace>2.0.co;2. Williams, E., et al. (2002), Contrasting convective regimes over the Amazon: Implications for cloud electrification, J. Geophys. Res., 107(D20), 8082, doi: /2001jd Yoshida, S., T. Morimoto, T. Ushio, and Z. Kawasaki (2007), ENSO and convective activities in Southeast Asia and western Pacific, Geophys. Res. Lett., 34, L21806, doi: /2007gl Yuan, T., and Z. Li (2010), General macro and microphysical properties of deep convective clouds as observed by MODIS, J. Clim., 23(13), , doi: /2009jcli Yuan, T., J. V. Martins, Z. Li, and L. A. Remer (2010), Estimating glaciation temperature of deep convective clouds with remote sensing data, Geophys. Res. Lett., 37, L08808, doi: /2010gl Zhang, R., G. Li, J. Fan, D. L. Wu, and M. J. Molina (2007), Intensification of Pacific storm track linked to Asian pollution, Proc. Natl. Acad. Sci. U. S. A., 104(13), , doi: /pnas Zipser, E. J. (1994), Deep cumulonimbus cloud systems in the tropics with and without lightning, Mon. Weather Rev., 122(8), , doi: / (1994)122<1837:dccsit>2.0.co;2. Zipser, E. J., and K. R. Lutz (1994), The vertical profile of radar reflectivity of convective cells: A strong indicator of storm intensity and lightning probability, Mon. Weather Rev., 122(8), , doi: / (1994)122<1751:TVPORR>2.0.CO;2. K. E. Pickering, Laboratory for Atmospheres, NASA Goddard Space Flight Center, Bldg. 33, Rm. E417, Mail Code 613.2, Greenbelt, MD 20771, USA. L. A. Remer, Laboratory for Atmospheres, NASA Goddard Space Flight Center, Bldg. 33, Rm. A313, Mail Code 613.2, Greenbelt, MD 20771, USA. H. Yu, Laboratory for Atmospheres, NASA Goddard Space Flight Center, Bldg. 33, Rm. C317, Mail Code 613.2, Greenbelt, MD 20771, USA. T. Yuan, Laboratory for Atmospheres, NASA Goddard Space Flight Center, Bldg. 33, Rm. A306, Mail Code 613.2, Greenbelt, MD 20771, USA. (tianle.yuan@nasa.gov) 5of5