Atmos. Chem. Phys., 21, 831–851, 2021 https://doi.org/10.5194/acp-21-831-2021 © Author(s) 2021. This work is distributed under the Creative Commons Attribution 4.0 License. Linking marine phytoplankton emissions, meteorological processes, and downwind particle properties with FLEXPART Kevin J. Sanchez1,2, Bo Zhang3, Hongyu Liu3, Georges Saliba4, Chia-Li Chen4, Savannah L. Lewis4, Lynn M. Russell4, Michael A. Shook2, Ewan C. Crosbie2,5, Luke D. Ziemba2, Matthew D. Brown2,5, Taylor J. Shingler2, Claire E. Robinson2,5, Elizabeth B. Wiggins1,2, Kenneth L. Thornhill2,5, Edward L. Winstead2,5, Carolyn Jordan2,3, Patricia K. Quinn6, Timothy S. Bates6,7, Jack Porter9, Thomas G. Bell8,9, Eric S. Saltzman9, Michael J. Behrenfeld10, and Richard H. Moore2 1NASA Postdoctoral Program, Universities Space Research Association, Columbia, MD, USA 2NASA Langley Research Center, Hampton, VA, USA 3National Institute of Aerospace, Hampton, VA, USA 4Scripps Institution of Oceanography, University of California San Diego, La Jolla, CA, USA 5Science Systems and Applications, Inc., Hampton, VA, USA 6Pacific Marine Environmental Laboratory, NOAA, Seattle, WA, USA 7Joint Institute for the Study of the Atmosphere and Ocean (JISAO), University of Washington, Seattle, WA, USA 8Plymouth Marine Laboratory, Prospect Place, Plymouth, United Kingdom 9The Department of Earth System Science, University of California, Irvine, CA, USA 10Department of Botany and Plant Pathology, Oregon State University, Corvallis, OR, USA Correspondence: Kevin J. Sanchez ([email protected]) and Richard H. Moore ([email protected]) Received: 11 July 2020 – Discussion started: 17 September 2020 Revised: 13 November 2020 – Accepted: 15 November 2020 – Published: 20 January 2021 Abstract. Marine biogenic particle contributions to at- lagged in time. In particular, the marine non-refractory or- mospheric aerosol concentrations are not well understood ganic aerosol mass correlates with modeled marine net pri- though they are important for determining cloud optical mary production when weighted by 5 d air mass trajectory and cloud-nucleating properties. Here we examine the rela- residence time (r D 0:62). This result indicates that non- tionship between marine aerosol measurements (with satel- refractory organic aerosol mass is influenced by biogenic lites and model fields of ocean biology) and meteorologi- volatile organic compound (VOC) emissions that are typ- cal variables during the North Atlantic Aerosols and Marine ically produced through bacterial degradation of dissolved Ecosystems Study (NAAMES). NAAMES consisted of four organic matter, zooplankton grazing on marine phytoplank- field campaigns between November 2015 and April 2018 ton, and as a by-product of photosynthesis by phytoplank- that aligned with the four major phases of the annual phy- ton stocks during advection into the region. This is further toplankton bloom cycle. The FLEXible PARTicle (FLEX- supported by the correlation of non-refractory organic mass PART) Lagrangian particle dispersion model is used to spa- with 2 d residence-time-weighted chlorophyll a (r D 0:39), tiotemporally connect these variables to ship-based aerosol a proxy for phytoplankton abundance, and 5 d residence- and dimethyl sulfide (DMS) observations. We find that cor- time-weighted downward shortwave forcing (r D 0:58), a relations between some aerosol measurements with satellite- requirement for photosynthesis. In contrast, DMS (formed measured and modeled variables increase with increasing through biological processes in the seawater) and primary trajectory length, indicating that biological and meteorolog- marine aerosol (PMA) concentrations showed better corre- ical processes over the air mass history are influential for lations with explanatory biological and meteorological vari- measured particle properties and that using only spatially ables weighted with shorter air mass residence times, which coincident data would miss correlative connections that are reflects their localized origin as primary emissions. Aerosol Published by Copernicus Publications on behalf of the European Geosciences Union. 832 K. J. Sanchez et al.: Linking marine phytoplankton emissions with FLEXPART submicron number and mass negatively correlate with sea in the winter months, increasing nutrients at the surface and surface wind speed. The negative correlation is attributed to decoupling phytoplankton from predators, followed by in- enhanced PMA concentrations under higher wind speed con- creased sunlight in the spring, enhancing photosynthetic pri- ditions. We hypothesized that the elevated total particle sur- mary productivity; together they produce the annually occur- face area associated with high PMA concentrations leads to ring North Atlantic phytoplankton bloom (Behrenfeld, 2010; enhanced rates of condensation of VOC oxidation products Behrenfeld and Boss, 2018; Boss and Behrenfeld, 2010). The onto PMA. Given the high deposition velocity of PMA rel- bloom ends when phytoplankton division rates stop increas- ative to submicron aerosol, PMA can limit the accumulation ing (due to depletion of nutrients or an annual maximum in of secondary aerosol mass. This study provides observational mixed layer light intensity) and are matched by loss rates evidence for connections between marine aerosols and un- (Behrenfeld and Boss, 2018). When bloom termination is as- derlying ocean biology through complex secondary forma- sociated with nutrient exhaustion, mixed later depths may tion processes, emphasizing the need to consider air mass continue to shoal into summer (i.e., mixed layer light lev- history in future analyses. els are still increasing), but phytoplankton biomass may de- crease due to slowing division rates and excessive grazing (Behrenfeld and Boss, 2018). Analysis of phytoplankton taxonomy and its seasonal vari- 1 Introduction ability in the NAAMES region is presented by Bolaños et al. (2020). Bolaños et al. (2020) show that cyanobacteria Marine environments are sensitive to aerosol particle loading dominated subpolar waters during the winter and were a sig- because particles can act as cloud condensation nuclei (CCN) nificant fraction in the subtropics, with taxa varying by lat- on which cloud droplets form. The number concentration itude. In addition, prasinophyta accounted for a significant of cloud droplets can influence cloud optical properties and contribution of subtropical species, with stramenopiles rep- therefore affect the impact of clouds on climate (Leahy et al., resenting less than 30 % of subtropical communities. Spring 2012; Platnick and Twomey, 1994; Turner et al., 2007; War- communities had significantly more diverse communities ren et al., 1988). Measurements over the ocean are scarce and and significantly less cyanobacteria (< 10 %) relative to the have historically been concentrated over relatively few areas winter, with the exception of one station. Prasinophyta dom- for short time periods aligning with episodic intensive ship or inated the spring phytoplankton composition, though taxo- aircraft campaigns. Satellite measurements are crucial to fill- nomic compositions differed from the winter period and be- ing this void in marine boundary layer (MBL) measurements, tween the subpolar and subtropical regions. Typically, di- but satellite optical measurements of particles are dispropor- atoms are assumed to be the dominant phytoplankton species tionally weighted by the largest optically active (> 300 nm in blooms. However, diatoms only represent 10 %–40 % of diameter) particles (Hasekamp et al., 2019). An unresolved phytoplankton biomass in the spring bloom surveyed dur- challenge is to relate these remote sensing measurements to ing NAAMES. The phytoplankton functional groups present the particle number, size distribution, and chemical compo- influence the overall isoprene production rate and therefore sition that are known to drive variability in cloud properties. the marine atmospheric aerosol and VOC concentrations. Since ocean-emitted volatile compounds and particles can Booge et al. (2016) compiled chlorophyll-normalized iso- control the number, size, and composition of marine aerosols prene production rates from the literature to identify dif- (Brooks and Thornton, 2018), here we use satellite measure- ferences between phytoplankton species. The chlorophyll- ments of ocean biomass as a proxy for marine particle prop- normalized isoprene production rates varied from 4.56–27.6, erties. 1.4–32.16, and 1.12–28.48 (µmol (g chlorophyll a)−1 d−1) In the absence of advected continental or anthropogenic for cyanobacteria, prasinophyta, and diatoms, respectively, pollution, the main source of MBL particles is sea salt indicating that emissions for isoprene vary significantly with from primary marine aerosol (PMA), dissolved organic mat- taxonomy. To further complicate the emission strength of ter, and marine biogenic volatile organic compound (VOC) VOCs, emissions can vary by production pathways, such emissions that can oxidize and condense to form secondary as photosynthetic by-products, bacterial degradation of dis- aerosol mass (Bates et al., 1998a; Covert et al., 1992; solved organic matter, and zooplankton grazing on marine Frossard et al., 2014b; Murphy et al., 1998; Quinn et al., phytoplankton (Gantt et al., 2009; Shaw et al., 2003; Sinha 2000, 2014; Rinaldi et al., 2010; Sievering et al., 1992, 1999; et al., 2007). One of the most studied biogenically produced 2− Warren and Seinfeld, 1985). Clear seasonal trends have been marine aerosol