Temporal-Scale Analysis of Environmental Controls on Sea 10.1029/2018GL078707 Spray Aerosol Production Over the South Paciﬁc Gyre

Geophysical Research Letters

RESEARCH LETTER Temporal-Scale Analysis of Environmental Controls on Sea 10.1029/2018GL078707 Spray Aerosol Production Over the South Paciﬁc Gyre

Key Points: Tom Dror1, Yoav Lehahn2, Orit Altaratz1 , and Ilan Koren1 • Time scale-dependent links are shown between sea spray aerosol 1Department of Earth and Planetary Sciences, Weizmann Institute of Science, Rehovot, Israel, 2Department of Marine amount and surface wind speed, with an anticorrelation on a seasonal Geosciences, Charney School of Marine Sciences, University of Haifa, Haifa, Israel scale • Seasonality in the link between sea spray amount and surface wind Abstract An understanding of sea spray aerosol (SSA) production is needed to better assess its influence speed indicates of additional factors on climate. Using satellite data, we investigated the production of the coarse mode of aerosol optical that impact sea spray aerosol fi production depth (AODc), a proxy for SSA, over the pristine South Paci c Gyre. The analysis was done on three time • Marine biological activity is scales: daily, seasonal, and interannual. Scale-dependent links were shown between the AODc and wind suggested to be a secondary speed (W). AODc and W were positively correlated on both daily and interannual time scales but were influence that suppresses sea spray fi aerosol production signi cantly anticorrelated on the seasonal time scale. Seasonality of the AODc W link suggests contribution of other environmental factors. The main variable that could statistically explain trends in AODc on the seasonal time scale was chlorophyll a concentration, which showed a clear negative correlation Supporting Information: • Supporting Information S1 with AODc. The AODc yield per W unit was clearly reduced when chlorophyll a concentration was high, suggesting a secondary, but important influence of marine biological activity on SSA production. Correspondence to: I. Koren, Plain Language Summary In this study we conduct temporal-scale analysis on satellite data to [email protected] explore the environmental factors that impact sea spray particles production. We focus on the South Pacific Ocean, which is a pristine marine environment and is characterized by low biological activity. The

Citation: production of sea spray particles, that is wind-driven process, is shown to be affected by the state of Dror, T., Lehahn, Y., Altaratz, O., & the marine biological activity (detected best on a seasonal time scale). High biological activity is suggested Koren, I. (2018). Temporal-scale analysis as a secondary inﬂuence that suppresses particles’ production. This discovery is important for estimates of environmental controls on sea spray ’ aerosol production over the South of marine particles budgets and effects and will enable better representation of marine aerosols in Paciﬁc Gyre. Geophysical Research climate models. Letters, 45, 8637–8646. https://doi.org/ 10.1029/2018GL078707

Received 10 MAY 2018 1. Introduction Accepted 10 AUG 2018 Accepted article online 20 AUG 2018 Marine aerosols—the aerosol forming in the marine atmospheric boundary layer (MBL)—are produced by Published online 24 AUG 2018 both primary and secondary processes. The primary marine aerosol, or sea spray aerosol (SSA), is produced through the interaction of wind with the ocean surface, resulting in the mechanical formation of aerosols (O’Dowd & De Leeuw, 2007). The secondary aerosol is produced from the conversion of available trace gas- eous molecules by chemical reactions into solid or liquid particles, in a gas-to-particle conversion mechanism (Larson, 1980; O’Dowd et al., 1997; Ramachandran, 2018). SSA is emitted from the surface of the ocean through bubble bursting (O’Dowd & De Leeuw, 2007). Ranging in size from submicrometer to a few micrometers (Lewis & Schwartz, 2004), it contains both inorganic (sea salts) and organic materials. SSA accounts for approximately 1 3×1016 g/year (Erickson & Duce, 1988; Gong, 2003) of the global aerosol mass, making it one of the major types of particulate matter in the atmosphere (Cochran et al., 2017; De Leeuw et al., 2011; Grini et al., 2002; Lewis & Schwartz, 2004). SSA often dominates the mass concentration of marine aerosol, especially in remote locations (De Leeuw et al., 2011; Fitzgerald, 1991). Consequently, the coarse mode (i.e., particles with radii>1 μm) of marine aero-

sol optical depth (AODc) is composed mainly of SSA (Fitzgerald, 1991; Kleefeld et al., 2002; Lewis & Schwartz, 2004; Sellegri et al., 2001). Different studies using various measurement methods have shown that the frac-

tional contribution of SSA to the marine AODc ranges between 50% and 100% (Bigg, 1980; Mészáros & Vissy, 1974; Quinn et al., 1998; Sellegri et al., 2001). Mcinnes et al. (1997) observed that in remote oceanic regions,

SSA accounts for 86–100% of supermicron aerosol particles. The ﬁne mode of marine AOD (AODf) can consist of some submicron SSA, but it is dominated by other types of materials (Lewis & Schwartz, 2004). Since about

©2018. American Geophysical Union. 98% of SSA mass is supermicron (Gong et al., 2002), together with the fact that in remote oceans, the marine All Rights Reserved. aerosol is dominated by SSA, we use the AODc as a proxy for SSA in the atmosphere.

DROR ET AL. 8637 Geophysical Research Letters 10.1029/2018GL078707

Being abundant particles, SSA have a critical impact on the radiation budget (Charlson & Schwartz, 1992; Murphy et al., 1998) and on the microphysical and macrophysical properties of clouds (Bender et al., 2016; Bretherton et al., 2004; Feingold et al., 2010; Fitzgerald, 1974; Fitzgerald & Spyers-Duran, 1973; Koren & Feingold, 2011; Ramanathan et al., 2001; Squires & Twomey, 1960; Twomey, 1977; Wood, 2012). Therefore, SSA is a key element in the climate system (O’Dowd & De Leeuw, 2007). Accurate representation of the factors and processes controlling SSA production, and the response of these formation processes to the changing marine environment (Xu et al., 2015), is therefore an important step toward precise modeling of climatic processes. Triggered by wind-driven processes, SSA formation has been shown to depend primarily on the surface wind speed (W; Ceburnis et al., 2008; Glantz et al., 2009; Gong, 2003; Korhonen et al., 2010; Lehahn et al., 2010; Monahan & Muircheartaigh, 1980; Mulcahy et al., 2008; Prijith et al., 2014; Smirnov et al., 2003), especially above a threshold of 4 m/s (Lehahn et al., 2010). Nevertheless, a number of studies have shown that the physical and chemical characteristics of the SSA population are influenced by the properties and content of the seawater (Grythe et al., 2014; Jaeglé et al., 2011; Kasparian et al., 2017; Ovadnevaite et al., 2014; Salter et al., 2015; Tyree et al., 2007). This applies mainly to the uppermost 1–1,000-μm layer of the ocean, which is defined as the sea surface microlayer (Engel et al., 2017; Hardy, 1982; Hunter & Liss, 1977). Jaeglé et al. (2011) showed a clear dependence of inorganic SSA emissions on sea surface temperature (SST) across multi- ple data sets, especially under conditions of high wind speed (W > 6 m/s). Grythe et al. (2014) compared existing SSA source functions with a global database of SSA mass concentration measurements and showed a strong influence of SST on SSA production. More recently, Salter et al. (2015) derived an inorganic SSA source function incorporating SST and employed it in large-scale models. Ovadnevaite et al. (2014) devel- oped a source function that included the Reynolds number, allowing to combine not only W as a forcing parameter but also the influence of wave height, wind history, friction velocity, and viscosity (which inher- ently includes the impact of SST and salinity) in one parameter. Kasparian et al. (2017) proposed an empirical parameterization of the organic-bearing particles’ concentration, which depends on water salinity and SST. Frossard et al. (2014) conducted a bubbling experiment and observed that the presence of a thick stable microlayer can prevent certain materials’ inclusion in the bubbles’ film and, consequently, might change the SSA composition. In recent years, the state of marine biological activity has also been shown to influence SSA properties (Cavalli et al., 2004; Fuentes et al., 2010a; Meskhidze & Nenes, 2010; O’Dowd et al., 2004, 2008, 2015; Sellegri et al., 2006; Tyree et al., 2007; Yoon et al., 2007). This follows from two decades of research that focused on the impact of biological activity on the formation of secondary aerosols following gas-to-particle conversion of the precursor dimethyl sulfide, emitted by phytoplankton (Charlson et al., 1987). Although extensive work has been invested in identifying the possible impact of biological activity on the properties of MBL aerosols (of both primary and secondary sources; Ayers & Gras, 1991; Korhonen et al., 2008; O’Dowd et al., 2004, 2015; Vallina et al., 2006), the nature of this link is still not clear (Quinn & Bates, 2011). Thus, although there is an overall agreement that seawater properties are likely to affect SSA production, observational-based quantification of SSA’s response to changes in seawater properties is still lacking, mainly due to (i) difficulty in distinguishing the contribution of locally produced SSA (to the aerosol population at a given location) from the contribution of aerosols transported from remote terrestrial or marine locations; and (ii) difficulty in untangling the secondary effect of changes in seawater properties from the dominant effect of changes in W. Here we address this challenge and estimate the effect of oceanic factors on SSA production with a focus on biological activity, using satellite observations over the South Pacific Subtropical Gyre (see boxes in Figure 1).

2. Data and Methods Our analysis used 8 years of satellite-retrieved data (2003–2010). To allow for simultaneous measurements, we used the products of different sensors onboard the same satellite (Aqua). The Aqua satellite (launched in May 2002) follows a sun-synchronous orbit, providing daily measurements with equatorial crossing times at 13:30 local solar time.

DROR ET AL. 8638 Geophysical Research Letters 10.1029/2018GL078707

Figure 1. Annual mean distribution of (a) coarse mode of aerosol optical depth, (b) wind speed, (c) chlorophyll a concentration, and (d) sea surface temperature over the oceans for the years 2003 to 2010. The red box marks the main region of interest, and the black boxes mark the additional examined regions of interest.

2.1. Satellite Data 2.1.1. Aerosol Properties Aerosol parameters were derived from the Moderate Resolution Imaging Spectroradiometer (MODIS) with spatial resolution of 1° × 1°. AOD and size parameters are standard MODIS overocean products, and even though the retrieval is considered to be less accurate for low AOD regimes, these products have been exten- sively validated (Levy et al., 2013; Remer et al., 2005). The data set consisted of daily and monthly level 3 AOD (retrieved at 550 nm) from the Collection 6 algorithm

of MODIS aerosol products (Levy et al., 2013). The MODIS aerosol products include the ﬁne mode fraction (ff), deﬁned as the ratio between AOD contributed by small mode particles (radii <1 μm) and the total AOD

(AODf/AODtot), and therefore reflect the contribution of fine mode particles to the total aerosol extinction (Remer et al., 2005). In this study, ff was used to calculate AODc using AODc =(1 ff) · AODtot. As discussed in section 1, AODc is a good measure for SSA amount in the MBL of the region of interest (ROI). 2.1.2. Signal Validation To validate MODIS’s AOD retrieval and ensure that the signal shows an actual increase (decrease) in aerosol amount and not changes in the background ocean color, we compared the MODIS AOD signal to the Visible Infrared Imaging Radiometer Suite (VIIRS) aerosol optical thickness retrieval for the year 2014. VIIRS is a 22-band cross-track scanning radiometer onboard the Suomi National Polar-Orbiting Partnership spacecraft (Cao et al., 2014). VIIRS measures reflected and emitted radiation from the Earth-atmosphere system in narrow discrete bands, similar to the MODIS instrument. VIIRS’s overocean AOD algorithm starts from the 672-nm band, but, unlike MODIS, the VIIRS instrument’s aerosol algorithm uses the green band (550 nm) only as part of internal screening tests over ocean and not to actually derive aerosol properties (Jackson et al., 2013). Thus, a change in ocean color should not affect the AOD retrieval. The comparison between MODIS and VIIRS showed good agreement throughout the entire year (see supporting information Figure S4), proving that MODIS’s AOD retrieval captures the changes in aerosols in the atmosphere and not the changes in the background ocean color. 2.1.3. Surface Wind Surface wind speeds (W, i.e., winds at 10-m height above sea level) were estimated from measurements taken by the Advanced Microwave Scanning Radiometer of NASA’s Earth Observing System sensor. The wind speed data were obtained from Remote Sensing Systems Inc. (http://www.remss.com/). The spatial resolution was 0.25° × 0.25°. Advanced Microwave Scanning Radiometer of NASA’s Earth Observing System is a passive

DROR ET AL. 8639 Geophysical Research Letters 10.1029/2018GL078707

Figure 2. Time series of daily AODc and W (thin blue and red lines in a, respectively); the darker solid lines represent smoothed values using a moving average filter of 90 days. Linear correlations between AODc and W for daily (b) and annual (c) time scales. The data were sorted according to W and divided into 15 bins of equal numbers of observations (see numbers inside the panels). The correlation coefficients and p values were calculated from the full data set before binning (b) and from the annual means (c). All correlations were found to be significant at over 99% confidence level (pval ≅ 0in both cases). AODc = coarse mode of aerosol optical depth; W = wind speed.

microwave radiometer that provides estimates of the wind speed. Estimates of wind speed from passive radiometers were validated against buoy data (Ebuchi et al., 2002; Mears et al., 2001) and, despite the differences in measuring methods, showed good agreement (Wentz et al., 2007). 2.1.4. Seawater Properties

Surface chlorophyll a (Chla) concentrations and SST were also derived from MODIS. The data set was composed of level 3 images obtained from theÀÁ ocean color data distribution site (http://oceandata.sci.gsfc. 1= 1= nasa.gov/). The spatial resolution was 9 km 12° 12° . 2.2. Region of Interest To inspect how ocean properties and W affect SSA loading in a controlled way, we looked for a region that was free of long-range transport of aerosol. Moreover, assuming that the marine productivity effect is mostly noted in transitions from extremity oligotrophic to slightly productive conditions, the selected main ROI covered a 10° × 10° area in the southern Paciﬁc Ocean between 22°S 32°S and 124°W 114°W (red box in Figure 1). The ROI is located 4,500 km from the nearest continent, away from known sources and transport routes of terrestrial aerosol, to minimize the contribution of anthropogenic, mineral dust, or wildﬁre particles.

The region is characterized by a mean annual AODc value of 0.089 ± 0.005 (Figure 1a) and AODf value of 0.0052 ± 7.8 · 10 4 (see Figure S5), which is among the lowest worldwide (Koren et al., 2014). Moreover, the ROI is characterized by low variance in meteorological conditions. This translates into small changes in the MBL height throughout the year (~1,000 ± 200 m). These environmental conditions and aerosol properties allow us to assume that particles emitted from the sea surface comprise most of the coarse mode aero-

sols in the MBL (Mcinnes et al., 1997) and, therefore, that AODc values provide a good measure for SSA. Assuming this, hereafter, we use AODc as a proxy for SSA abundance over the ROI. The relatively low AODc values are associated with a moderate annual mean W (7.34 ± 0.93 m/s; Figure 1b). From an oceanic perspec- tive, the region is characterized by very low levels of biological productivity, largely due to limited availability of nutrients in the upper, well-lit surface layer (Claustre & Maritorena, 2003; Dandonneau et al., 2003; Polovina 3 et al., 2008). This is reﬂected in low annual mean values of Chla (0.027 ± 0.012 mg/m ; Figure 1c) in a region of

DROR ET AL. 8640 Geophysical Research Letters 10.1029/2018GL078707

Figure 3. Daily means (each point represents the average value of the same day of the year for 8 years of data, 2003 to 2010). The thick lines show smoothed values (using a moving average filter of 10 days) and standard deviations (shaded areas) of (a) AODc, (b) W, (c) Chla concentration, and (d) SST. Scatter plots and correlations between AODc and (e) W, (f) Chla concentration, and (g) SST. The data were sorted according to W, Chla, and SST (e–g, respectively) and divided into 15 bins of equal numbers of observations. The correlation coefficients and p values were calculated from the full data set before binning. All correlations were found to be significant at over 99% confidence level (pval ≅ 0 in all cases). AODc = coarse mode of aerosol optical depth; Chla = chlorophyll a; W = wind speed; SST = sea surface temperature.

relatively high SST (22.27 ± 1.85°C; Figure 1d). The phytoplankton bloom in this region peaks in austral winter (Figure 3c). In addition to the main ROI, similar analysis was done for two neighboring 10° × 10° boxes that span alto- gether a ~3,000-km belt over the Pacific Ocean. This belt is located south of the Tropical Pacific, and therefore, the interannual influence of the El Niño-Southern Oscillation is less pronounced (Xu et al., 2015). 2.3. Statistical Analysis

Daily data of aerosol parameters, W, Chla level, and SST, were extracted over a 10° × 10° ROI and averaged for the required time period. The ROI was covered by 100 pixels, where W, Chla, and SST data were averaged from ﬁ 1= 1= ðÞ a ner resolution of 0.25° × 0.25° (W) and 12° 12°9km(Chla and SST) into 1° × 1° resolution to match the AOD data. The data were sorted according to W, Chla, and SST and divided into 15 bins of equal numbers of samples. Slope values, correlation coefﬁcients, and p values (using t test) were calculated from the full data set before binning.

3. Results We studied the effects of varying environmental conditions on SSA production over the South Paciﬁc

Subtropical Gyre (see red and black boxes in Figure 1). Temporal variations and mean values of AODc and

DROR ET AL. 8641 Geophysical Research Letters 10.1029/2018GL078707

their association to changes in environmental conditions were analyzed over three time scales: daily, seasonal, and interannual. On a daily time

scale, the AODc signal was highly variable, with abrupt day-to-day changes that ranged between 0.006% and 55% of the annual mean (thin blue line in Figure 2a). In agreement with previous studies (Glantz et al., 2009; Korhonen et al., 2010; Lehahn et al., 2010; Monahan & Muircheartaigh, 1980; Mulcahy et al., 2008; Prijith et al., 2014; Smirnov et al., 2003), these

high-frequency variations in AODc were signiﬁcantly correlated with W (Figure 2b; see p values in the ﬁgure caption), which was also characterized by high variability (thin red line in Figure 2a). In contrast, a comparison of

the seasonal cycles of AODc and W signals derived by applying a 90-day moving average window (thick blue and red lines in Figure 2a, respectively) showed no clear dependency: the two smoothed time series were characterized by distinctly different seasonal cycles. Finally, comparing

Figure 4. AODc as a function of wind speed for the mean of JJA (austral win- annual averages of AODc and W showed a signiﬁcant correlation ter, blue) and DJF (austral summer, red) for the years 2003–2010. The data (Figure 2c), indicating that the observed interannual variability in AODc were sorted according to wind speed and divided into 15 bins of equal ﬁ numbers of observations (n = 2,670 and n = 1,496 for JJA and DJF, respec- can be explained, to a rst order, by year-to-year variations in the mean tively). Points and shaded areas represent the mean and standard deviation W. The apparent differences in the smoothed AODc and W signals (thick of AODc for each bin. The solid lines represent the linear regression that was blue and red lines in Figure 2a, respectively) suggest that on a seasonal performed from the full cloud of data points above a wind speed time scale, the variability in AODc over the ROI is impacted by changes threshold of 4 m/s (Lehahn et al., 2010). AODc = coarse mode of aerosol in other environmental factors. optical depth; JJA = June-July-August; DJF = December-January-February. To further investigate this effect, we compared the seasonal cycles of different parameters as calculated for 8 years of data for the main ROI (Figures 3a–3d) and for the two additional

ROIs (Figures S1 and S2). AODc and W were characterized by opposite seasonal cycles, with an annual minimum in AODc during austral winter (June-July-August; Figure 3a) that was associated with an annual maximum in W (Figure 3b). A similar picture emerged for the case of phytoplankton biomass as expressed by

Chla concentration, with the low wintertime AODc levels being associated with an annual maximum in Chla levels (Figure 3c). These high Chla levels represent the wintertime phytoplankton bloom, which results from the seasonal deepening of the oceanic-mixed layer. The SST was characterized by a different seasonal cycle, with maximum and minimum values observed during the austral fall and spring, respectively (Figure 3d). These results were consistent for the three ROIs. The relations between the seasonal cycles of the different parameters (as shown in Figures 3a–3d, S1, and S2) were reexamined in scatter plots of mean daily data (Figures 3f and 3g), where each of the 365 points represents an average of the same day (along the year) over 8 years. Signiﬁcant correlations were found between

the mean daily signals of AODc and W (R = 0.15), between AODc and Chla (R = 0.57; this was done also for the two additional ROIs, see Figure S3), and between AODc and SST (R = 0.035; Figures 3f and 3g, respectively). Although signiﬁcant, the low R and slope values showed that on a seasonal time scale, changes in

both W and SST cannot explain the AODc signal. Finally, to further highlight the emerging link between SSA production and biological activity that changes signiﬁcantly on a seasonal time scale, we compared two seasons. We examined the relations between daily

time series of AODc and W during austral winter when AODc is minimal and Chla is at its maximal level and the austral summer (December-January-February) when AODc is maximal and Chla is minimal (blue and red lines in Figure 4, respectively).

This analysis isolated, to a first approximation, the effect of biological activity (measured by Chla level) on SSA production from the W effect. It showed significant differences in the dependency of AODc on W between the two seasons. The regression slopes between the two fields were 0.005 in austral winter when the biological production is at its peak (Figure 3c) and 0.011 in the summer when biological production is minimal.

4. Discussion and Conclusions Using satellite observations of atmospheric and oceanic variables, we assessed the effect of environmental conditions on the production of SSA in a pristine region over the South Paciﬁc Subtropical Gyre. We found

DROR ET AL. 8642 Geophysical Research Letters 10.1029/2018GL078707

that the association between SSA (which was approximated by satellite-derived AODc) and some environmental factors (including W, SST, and Chla) depends on the time scale under consideration. While the wind is the crucial mechanism that produces SSA, we speciﬁcally distinguished between two time regimes: (i) a

dominant wind regime, in which the AODc trend can be explained by changes in W. This regime emerges for the daily and interannual time scales; and (ii) a regime in which AODc variability cannot be fully explained by W but by other environmental oceanic factors. This regime matched the seasonal time scale.

The observed day-to-day dependence of AODc on W is in agreement with many previous studies (Glantz et al., 2009; Korhonen et al., 2010; Lehahn et al., 2010; Monahan & Muircheartaigh, 1980; Mulcahy et al., 2008; Smirnov et al., 2003), which all showed the strong dependence of SSA emission on surface wind speed, the driving force for its production (Lewis & Schwartz, 2004). Our results showed that this dependency also translates to the interannual time scale, such that year-to-year changes in mean W drive changes in annually averaged SSA production. This result suggests that even a small increase in W, as expected for a warmer climate (Flohn & Kapala, 1989; Thomas et al., 2008; Young et al., 2011), will lead to enhanced SSA production and, consequently, to an increased amount of marine aerosols. In contrast to the observed dependence on short and long (i.e., daily and interannual, respectively) time

scales on the intermediate (seasonal) time scale, AODc was negatively correlated with W. This suggests that on such time scale, another factor may inﬂuence SSA production more signiﬁcantly than W (see Figures S6

and S7). Based on the satellite-derived annual trend of Chla, which was anticorrelated with that of AODc (Figure 3f), our results suggest that SSA production as examined on a seasonal time scale is partly associated with the change in marine biological activity, which controls the properties of the microlayer. A similar

link between AODc and Chla exists also on the interannual time scale (see Figure S8). Changes in microlayer properties have been shown to affect SSA characteristics (Engel et al., 2017; Frossard et al., 2014; Gantt et al., 2011). The microlayer’s composition determines the water surface tension, which in turn modulates fluxes to the atmosphere (Elliott et al., 2017). Specifically, our results suggest that the wintertime increase in phytoplankton biomass in the highly oligotrophic water of the South Pacific Subtropical Gyre suppresses SSA production, resulting in relatively low

levels of AODc. Anticorrelation between biological activity and SSA abundance was recently observed (Lehahn et al., 2014), based on in situ measurements in the North Atlantic. This hypothesis of enhanced biological activity resulting in reduced SSA production agrees with previous laboratory experiments that proposed an effect of high surfactant amounts in the surface microlayer (which are associated with high biological activity conditions) on the primary production of marine aerosols (Fuentes et al., 2010b; Modini et al., 2013; Sellegri et al., 2006). Modini et al. (2013) reported that increased amounts of surfactants in the microlayer decreased the efficiency of aerosol production by 79–98%. The suggested mechanism involved the influence of surfactants on bubble stability: addition of soluble surfactants to the surface seawater increased bubble persistence, leading to thinning of the bubble film, which may result in reduced aerosol production. As the main driver of SSA production, untangling the W effect on SSA production is critically important and

challenging. The fact that the link between W and AODc—which serves as a good proxy for SSA concentration—is dependent on season provides a strong indication that indeed, this link is affected by environmental factors that change the mixed layer, and more importantly, the oceanic microlayer’s properties, and therefore regulate the wind-driven SSA production. More speciﬁcally, the analysis of separate time scales and the daily analysis per season provided a measure of the secondary albeit important role of phytoplankton in the process, improving our quantitative understanding of the factors that control the amount of SSA in the Acknowledgments oceanic environment. This work was supported by a research grant from Scott Jordan and Gina Valdez. All data used in this study are publicly available and can be found on: https://ladsweb.modaps.eosdis.nasa. References gov/ (aerosol data), http://oceandata. Ayers, G., & Gras, J. (1991). Seasonal relationship between cloud condensation nuclei and aerosol methanesulphonate in marine air. Nature, sci.gsfc.nasa.gov/ (ocean properties 353(6347), 834–835. https://doi.org/10.1038/353834a0 data), and http://www.remss.com/ (sur- Bender, F. A.-M., Engström, A., & Karlsson, J. (2016). Factors controlling cloud albedo in marine subtropical stratocumulus regions in climate face wind data). All other data used models and satellite observations. Journal of Climate, 29(10), 3559–3587. https://doi.org/10.1175/JCLI-D-15-0095.1 have been cited, with details provided Bigg, E. (1980). Comparison of aerosol at four baseline atmospheric monitoring stations. Journal of Applied Meteorology, 19(5), 521–533. in references. https://doi.org/10.1175/1520-0450(1980)019<0521:COAAFB>2.0.CO;2

DROR ET AL. 8643 Geophysical Research Letters 10.1029/2018GL078707

Bretherton, C. S., Uttal, T., Fairall, C. W., Yuter, S. E., Weller, R. A., Baumgardner, D., et al. (2004). The EPIC 2001 stratocumulus study. Bulletin of the American Meteorological Society, 85(7), 967–978. https://doi.org/10.1175/BAMS-85-7-967 Cao, C., De Luccia, F. J., Xiong, X., Wolfe, R., & Weng, F. (2014). Early on-orbit performance of the visible infrared imaging radiometer suite onboard the Suomi National Polar-Orbiting Partnership (S-NPP) satellite. IEEE Transactions on Geoscience and Remote Sensing, 52(2), 1142–1156. https://doi.org/10.1109/TGRS.2013.2247768 Cavalli, F., Facchini, M., Decesari, S., Mircea, M., Emblico, L., Fuzzi, S., et al. (2004). Advances in characterization of size-resolved organic matter in marine aerosol over the North Atlantic. Journal of Geophysical Research, 109, D24215. https://doi.org/10.1029/2004JD005137 Ceburnis, D., O’Dowd, C. D., Jennings, G. S., Facchini, M. C., Emblico, L., Decesari, S., et al. (2008). Marine aerosol chemistry gradients: Elucidating primary and secondary processes and fluxes. Geophysical Research Letters, 35, L07804. https://doi.org/10.1029/2008GL033462 Charlson, R. J., Lovelock, J. E., Andreae, M. O., & Warren, S. G. (1987). Oceanic phytoplankton, atmospheric sulphur, cloud albedo and climate. Nature, 326(6114), 655–661. Charlson, R. J., & Schwartz, S. (1992). Climate forcing by anthropogenic aerosols. Science, 255(5043), 423–430. https://doi.org/10.1126/ science.255.5043.423 Claustre, H., & Maritorena, S. (2003). The many shades of ocean blue. Science, 302(5650), 1514–1515. https://doi.org/10.1126/science.1092704 Cochran, R. E., Ryder, O. S., Grassian, V. H., & Prather, K. A. (2017). Sea spray aerosol: The chemical link between the oceans, atmosphere, and climate. Accounts of Chemical Research, 50(3), 599–604. https://doi.org/10.1021/acs.accounts.6b00603 Dandonneau, Y., Vega, A., Loisel, H., Du Penhoat, Y., & Menkes, C. (2003). Oceanic Rossby waves acting as a “hay rake” for ecosystem floating by-products. Science, 302(5650), 1548–1551. https://doi.org/10.1126/science.1090729 De Leeuw, G., Andreas, E. L., Anguelova, M. D., Fairall, C., Lewis, E. R., O’Dowd, C., et al. (2011). Production flux of sea spray aerosol. Reviews of Geophysics, 49, RG2001. https://doi.org/10.1029/2010RG000349 Ebuchi, N., Graber, H. C., & Caruso, M. J. (2002). Evaluation of wind vectors observed by QuikSCAT/SeaWinds using ocean buoy data. Journal of Atmospheric and Oceanic Technology, 19(12), 2049–2062. https://doi.org/10.1175/1520-0426(2002)019<2049:EOWVOB>2.0.CO;2 Elliott, S., Burrows, S., Cameron-Smith, P., Hoffman, F., Hunke, E., Jeffery, N., et al. (2017). Does marine surface tension have global biogeo- graphy? Addition for the OCEANFILMS Package. Engel, A., Bange, H. W., Cunliffe, M., Burrows, S. M., Friedrichs, G., Galgani, L., et al. (2017). The ocean’s vital skin: Toward an integrated understanding of the sea surface microlayer. Frontiers in Marine Science, 4, 165. Erickson, D. J., & Duce, R. A. (1988). On the global flux of atmospheric sea salt. Journal of Geophysical Research, 93(C11), 14,079–14,088. https://doi.org/10.1029/JC093iC11p14079 Feingold, G., Koren, I., Wang, H., Xue, H., & Brewer, W. A. (2010). Precipitation-generated oscillations in open cellular cloud fields. Nature, 466(7308), 849–852. https://doi.org/10.1038/nature09314 Fitzgerald, J. W. (1974). Effect of aerosol composition on cloud droplet size distribution: A numerical study. Journal of the Atmospheric Sciences, 31(5), 1358–1367. https://doi.org/10.1175/1520-0469(1974)031<1358:EOACOC>2.0.CO;2 Fitzgerald, J. W. (1991). Marine aerosols: A review. Atmospheric Environment. Part A. General Topics, 25(3–4), 533–545. https://doi.org/10.1016/ 0960-1686(91)90050-H Fitzgerald, J. W., & Spyers-Duran, P. (1973). Changes in cloud nucleus concentration and cloud droplet size distribution associated with pollution from St. Louis. Journal of Applied Meteorology, 12(3), 511–516. https://doi.org/10.1175/1520-0450(1973)012<0511: CICNCA>2.0.CO;2 Flohn, H., & Kapala, A. (1989). Changes of tropical sea-air interaction processes over a 30-year period. Nature, 338(6212), 244–246. https://doi. org/10.1038/338244a0 Frossard, A. A., Russell, L. M., Burrows, S. M., Elliott, S. M., Bates, T. S., & Quinn, P. K. (2014). Sources and composition of submicron organic mass in marine aerosol particles. Journal of Geophysical Research: Atmospheres, 119, 12,977–13,003. https://doi.org/10.1002/ 2014JD021913 Fuentes, E., Coe, H., Green, D., Leeuw, G. d., & McFiggans, G. (2010a). Laboratory-generated primary marine aerosol via bubble-bursting and atomization. Atmospheric Measurement Techniques, 3(1), 141–162. https://doi.org/10.5194/amt-3-141-2010 Fuentes, E., Coe, H., Green, D., Leeuw, G. d., & McFiggans, G. (2010b). On the impacts of phytoplankton-derived organic matter on the properties of the primary marine aerosol—Part 1: Source fluxes. Atmospheric Chemistry and Physics, 10(19), 9295–9317. https://doi.org/ 10.5194/acp-10-9295-2010 Gantt, B., Meskhidze, N., Facchini, M., Rinaldi, M., Ceburnis, D., & O’Dowd, C. (2011). Wind speed dependent size-resolved parameterization for the organic mass fraction of sea spray aerosol. Atmospheric Chemistry and Physics, 11(16), 8777–8790. https://doi.org/10.5194/acp-11- 8777-2011 Glantz, P., Nilsson, E., & von Hoyningen-Huene, W. (2009). Estimating a relationship between aerosol optical thickness and surface wind speed over the ocean. Atmospheric Research, 92(1), 58–68. https://doi.org/10.1016/j.atmosres.2008.08.010 Gong, S. (2003). A parameterization of sea-salt aerosol source function for sub- and super-micron particles. Global Biogeochemical Cycles, 17(4), 1097. https://doi.org/10.1029/2003GB002079 Gong, S. L., Barrie, L. A., & Lazare, M. (2002). Canadian Aerosol Module (CAM): A size-segregated simulation of atmospheric aerosol processes for climate and air quality models 2. Global sea-salt aerosol and its budgets. Journal of Geophysical Research, 107(D24), 4779. https://doi. org/10.1029/2001JD002004 Grini, A., Myhre, G., Sundet, J. K., & Isaksen, I. S. (2002). Modeling the annual cycle of sea salt in the global 3D model Oslo CTM2: Concentrations, fluxes, and radiative impact. Journal of Climate, 15(13), 1717–1730. https://doi.org/10.1175/1520-0442(2002)015<1717: MTACOS>2.0.CO;2 Grythe, H., Ström, J., Krejci, R., Quinn, P., & Stohl, A. (2014). A review of sea-spray aerosol source functions using a large global set of sea salt aerosol concentration measurements. Atmospheric Chemistry and Physics, 14(3), 1277–1297. https://doi.org/10.5194/acp-14-1277-2014 Hardy, J. T. (1982). The sea surface microlayer: Biology, chemistry and anthropogenic enrichment. Progress in Oceanography, 11(4), 307–328. https://doi.org/10.1016/0079-6611(82)90001-5 Hunter, K., & Liss, P. (1977). The input of organic material to the oceans: Air—sea interactions and the organic chemical composition of the sea surface. Marine Chemistry, 5(4–6), 361–379. https://doi.org/10.1016/0304-4203(77)90029-9 Jackson, J. M., Liu, H., Laszlo, I., Kondragunta, S., Remer, L. A., Huang, J., & Huang, H. C. (2013). Suomi-NPP VIIRS aerosol algorithms and data products. Journal of Geophysical Research: Atmospheres, 118, 12,673–12,689. https://doi.org/10.1002/2013JD020449 Jaeglé, L., Quinn, P., Bates, T., Alexander, B., & Lin, J.-T. (2011). Global distribution of sea salt aerosols: New constraints from in situ and remote sensing observations. Atmospheric Chemistry and Physics, 11(7), 3137–3157. https://doi.org/10.5194/acp-11-3137-2011 Kasparian, J., Hassler, C., Ibelings, B., Berti, N., Bigorre, S., Djambazova, V., et al. (2017). Assessing the dynamics of organic aerosols over the North Atlantic Ocean. Scientific Reports, 7(1), 45476. https://doi.org/10.1038/srep45476

DROR ET AL. 8644 Geophysical Research Letters 10.1029/2018GL078707

Kleefeld, C., O’Dowd, C. D., O’Reilly, S., Jennings, S. G., Aalto, P., Becker, E., et al. (2002). Relative contribution of submicron and supermicron particles to aerosol light scattering in the marine boundary layer. Journal of Geophysical Research, 107(D19), 8103. https://doi.org/10.1029/ 2000JD000262 Koren, I., Dagan, G., & Altaratz, O. (2014). From aerosol-limited to invigoration of warm convective clouds. Science, 344(6188), 1143–1146. https://doi.org/10.1126/science.1252595 Koren, I., & Feingold, G. (2011). Aerosol–cloud–precipitation system as a predator-prey problem. Proceedings of the National Academy of Sciences, 108(30), 12,227–12,232. https://doi.org/10.1073/pnas.1101777108 Korhonen, H., Carslaw, K. S., Forster, P. M., Mikkonen, S., Gordon, N. D., & Kokkola, H. (2010). Aerosol climate feedback due to decadal increases in Southern Hemisphere wind speeds. Geophysical Research Letters, 37, L02805. https://doi.org/10.1029/ 2009GL041320 Korhonen, H., Carslaw, K. S., Spracklen, D. V., Mann, G. W., & Woodhouse, M. T. (2008). Influence of oceanic dimethyl sulfide emissions on cloud condensation nuclei concentrations and seasonality over the remote Southern Hemisphere oceans: A global model study. Journal of Geophysical Research, 113, D15204. https://doi.org/10.1029/2007JD009718 Larson, T. V. (1980). Secondary aerosol: Production mechanisms of sulfate compounds in the atmosphere. Annals of the New York Academy of Sciences, 338(1), 26–38. Lehahn, Y., Koren, I., Boss, E., Ben-Ami, Y., & Altaratz, O. (2010). Estimating the maritime component of aerosol optical depth and its dependency on surface wind speed using satellite data. Atmospheric Chemistry and Physics, 10(14), 6711–6720. https://doi.org/10.5194/ acp-10-6711-2010 Lehahn, Y., Koren, I., Rudich, Y., Bidle, K. D., Trainic, M., Flores, J. M., et al. (2014). Decoupling atmospheric and oceanic factors affecting aerosol loading over a cluster of mesoscale North Atlantic eddies. Geophysical Research Letters, 41, 4075–4081. https://doi.org/10.1002/ 2014GL059738 Levy, R., Mattoo, S., Munchak, L., Remer, L., Sayer, A., Patadia, F., & Hsu, N. (2013). The Collection 6 MODIS aerosol products over land and ocean. Atmospheric Measurement Techniques, 6(11), 2989–3034. https://doi.org/10.5194/amt-6-2989-2013 Lewis, E. R., & Schwartz, S. E. (2004). Sea salt aerosol production: Mechanisms, methods, measurements, and models—A critical review. Washington, DC: American Geophysical Union. Mcinnes, L., Covert, D., & Baker, B. (1997). The number of sea-salt, sulfate, and carbonaceous particles in the marine atmosphere. Tellus B, 49(3), 300–313. https://doi.org/10.3402/tellusb.v49i3.15969 Mears, C., Smith, D. K., & Wentz, F. J. (2001). Comparison of Special Sensor Microwave Imager and buoy-measured wind speeds from 1987 to 1997. Journal of Geophysical Research, 106(C6), 11,719–11,729. https://doi.org/10.1029/1999JC000097 Meskhidze, N., & Nenes, A. (2010). Effects of ocean ecosystem on marine aerosol-cloud interaction. Advances in Meteorology, 2010, 239808. https://doi.org/10.1155/2010/239808 Mészáros, A., & Vissy, K. (1974). Concentration, size distribution and chemical nature of atmospheric aerosol particles in remote oceanic areas. Journal of Aerosol Science, 5(1), 101–109. https://doi.org/10.1016/0021-8502(74)90011-1 Modini, R., Russell, L., Deane, G., & Stokes, M. (2013). Effect of soluble surfactant on bubble persistence and bubble-produced aerosol particles. Journal of Geophysical Research: Atmospheres, 118, 1388–1400. https://doi.org/10.1002/jgrd.50186 Monahan, E. C., & Muircheartaigh, I. (1980). Optimal power-law description of oceanic whitecap coverage dependence on wind speed. Journal of Physical Oceanography, 10(12), 2094–2099. https://doi.org/10.1175/1520-0485(1980)010<2094:OPLDOO>2.0.CO;2 Mulcahy, J., O’Dowd, C., Jennings, S., & Ceburnis, D. (2008). Significant enhancement of aerosol optical depth in marine air under high wind conditions. Geophysical Research Letters, 35, L16810. https://doi.org/10.1029/2008GL034303 Murphy, D., Anderson, J., Quinn, P., McInnes, L., Brechtel, F., Kreidenweis, S., et al. (1998). Influence of sea-salt on aerosol radiative properties in the Southern Ocean marine boundary layer. Nature, 392(6671), 62–65. https://doi.org/10.1038/32138 O’Dowd, C., Ceburnis, D., Ovadnevaite, J., Bialek, J., Stengel, D. B., Zacharias, M., et al. (2015). Connecting marine productivity to sea-spray via nanoscale biological processes: Phytoplankton dance or death disco? Scientific Reports, 5, 14883. O’Dowd, C., & De Leeuw, G. (2007). Marine aerosol production: A review of the current knowledge. Philosophical Transactions of the Royal Society of London A: Mathematical, Physical and Engineering Sciences, 365(1856), 1753–1774. O’Dowd, C. D., Facchini, M. C., Cavalli, F., Ceburnis, D., Mircea, M., Decesari, S., et al. (2004). Biogenically driven organic contribution to marine aerosol. Nature, 431(7009), 676–680. https://doi.org/10.1038/nature02959 O’Dowd, C. D., Langmann, B., Varghese, S., Scannell, C., Ceburnis, D., & Facchini, M. C. (2008). A combined organic-inorganic sea-spray source function. Geophysical Research Letters, 35, L01801. https://doi.org/10.1029/2007GL030331 O’Dowd, C. D., Smith, M. H., Consterdine, I. E., & Lowe, J. A. (1997). Marine aerosol, sea-salt, and the marine sulphur cycle: A short review. Atmospheric Environment, 31(1), 73–80. https://doi.org/10.1016/S1352-2310(96)00106-9 Ovadnevaite, J., Manders, A., de Leeuw, G., Ceburnis, D., Monahan, C., Partanen, A. I., et al. (2014). A sea spray aerosol flux parameterization encapsulating wave state. Atmospheric Chemistry and Physics, 14(4), 1837–1852. https://doi.org/10.5194/acp-14-1837-2014 Polovina, J. J., Howell, E. A., & Abecassis, M. (2008). Ocean’s least productive waters are expanding. Geophysical Research Letters, 35, L03618. https://doi.org/10.1029/2007GL031745 Prijith, S., Aloysius, M., & Mohan, M. (2014). Relationship between wind speed and sea salt aerosol production: A new approach. Journal of Atmospheric and Solar-Terrestrial Physics, 108,34–40. https://doi.org/10.1016/j.jastp.2013.12.009 Quinn, P., & Bates, T. (2011). The case against climate regulation via oceanic phytoplankton sulphur emissions. Nature, 480(7375), 51–56. https://doi.org/10.1038/nature10580 Quinn, P., Coffman, D., Kapustin, V., Bates, T., & Covert, D. (1998). Aerosol optical properties in the marine boundary layer during the First Aerosol Characterization Experiment (ACE 1) and the underlying chemical and physical aerosol properties. Journal of Geophysical Research, 103(D13), 16,547–16,563. https://doi.org/10.1029/97JD02345 Ramachandran, S. (2018). Aerosols and climate change: present understanding, challenges, and future outlook. In K. P. Vadrevu, T. Ohara, & C. Justice (Eds.), Land-Atmospheric Research Applications in South and Southeast Asia (pp. 341–378). Cham, Switzerland: Springer International Publishing. https://doi.org/10.1007/978-3-319-67474-2_17 Ramanathan, V., Crutzen, P. J., Kiehl, J. T., & Rosenfeld, D. (2001). Aerosols, climate, and the hydrological cycle. Science, 294(5549), 2119–2124. https://doi.org/10.1126/science.1064034 Remer, L. A., Kaufman, Y., Tanré, D., Mattoo, S., Chu, D., Martins, J. V., et al. (2005). The MODIS aerosol algorithm, products, and validation. Journal of the Atmospheric Sciences, 62(4), 947–973. https://doi.org/10.1175/JAS3385.1 Salter, M. E., Zieger, P., Acosta Navarro, J. C., Grythe, H., Kirkevåg, A., Rosati, B., et al. (2015). An empirically derived inorganic sea spray source function incorporating sea surface temperature. Atmospheric Chemistry and Physics, 15(19), 11,047–11,066. https://doi.org/10.5194/acp- 15-11047-2015

DROR ET AL. 8645 Geophysical Research Letters 10.1029/2018GL078707

Sellegri, K., Gourdeau, J., Putaud, J. P., & Despiau, S. (2001). Chemical composition of marine aerosol in a Mediterranean coastal zone during the FETCH experiment. Journal of Geophysical Research, 106(D11), 12,023–12,037. Sellegri, K., O’Dowd, C., Yoon, Y., Jennings, S., & de Leeuw, G. (2006). Surfactants and submicron sea spray generation. Journal of Geophysical Research, 111, D22215. https://doi.org/10.1029/2005JD006658 Smirnov, A., Holben, B., Eck, T., Dubovik, O., & Slutsker, I. (2003). Effect of wind speed on columnar aerosol optical properties at Midway Island. Journal of Geophysical Research, 108(D24), 4802. https://doi.org/10.1029/2003JD003879 Squires, P., & Twomey, S. (1960). The relation between cloud droplet spectra and the spectrum of cloud nuclei. Paper presented at the Physics of Precipitation: Proceedings of the Cloud Physics Conference, Woods Hole, Massachusetts, June 3-5, 1959. Thomas, B. R., Kent, E. C., Swail, V. R., & Berry, D. I. (2008). Trends in ship wind speeds adjusted for observation method and height. International Journal of Climatology, 28(6), 747–763. https://doi.org/10.1002/joc.1570 Twomey, S. (1977). The inﬂuence of pollution on the shortwave albedo of clouds. Journal of the Atmospheric Sciences, 34(7), 1149–1152. https://doi.org/10.1175/1520-0469(1977)034<1149:TIOPOT>2.0.CO;2 Tyree, C. A., Hellion, V. M., Alexandrova, O. A., & Allen, J. O. (2007). Foam droplets generated from natural and artiﬁcial seawaters. Journal of Geophysical Research, 112, D12204. https://doi.org/10.1029/2006JD007729 Vallina, S. M., Simó, R., & Gassó, S. (2006). What controls CCN seasonality in the Southern Ocean? A statistical analysis based on satellite- derived chlorophyll and CCN and model-estimated OH radical and rainfall. Global Biogeochemical Cycles, 20, GB1014. https://doi.org/ 10.1029/2005GB002597 Wentz, F. J., Ricciardulli, L., Hilburn, K., & Mears, C. (2007). How much more rain will global warming bring? Science, 317(5835), 233–235. https://doi.org/10.1126/science.1140746 Wood, R. (2012). Stratocumulus clouds. Monthly Weather Review, 140(8), 2373–2423. https://doi.org/10.1175/MWR-D-11-00121.1 Xu, L., Pierce, D. W., Russell, L. M., Miller, A. J., Somerville, R. C., Twohy, C. H., et al. (2015). Interannual to decadal climate variability of sea salt aerosols in the coupled climate model CESM1.0. Journal of Geophysical Research: Atmospheres, 120, 1502–1519. https://doi.org/10.1002/ 2014JD022888 Yoon, Y., Ceburnis, D., Cavalli, F., Jourdan, O., Putaud, J., Facchini, M., et al. (2007). Seasonal characteristics of the physicochemical properties of North Atlantic marine atmospheric aerosols. Journal of Geophysical Research, 112, D04206. https://doi.org/10.1029/2005JD007044 Young, I., Zieger, S., & Babanin, A. V. (2011). Global trends in wind speed and wave height. Science, 332(6028), 451–455. https://doi.org/ 10.1126/science.1197219

DROR ET AL. 8646