Assessing the Dynamics of Organic Aerosols over the North Atlantic Ocean

The influence of aerosols on climate is highly dependent on the particle size distribution, concentration, and composition. In particular, the latter influences their ability to act as cloud condensation nuclei, whereby they impact cloud coverage and precipitation. Here, we simultaneously measured the concentration of aerosols from sea spray over the North Atlantic on board the exhaust-free solar-powered vessel “PlanetSolar”, and the sea surface physico-chemical parameters. We identified organic-bearing particles based on individual particle fluorescence spectra. Organic-bearing aerosols display specific spatio-temporal distributions as compared to total aerosols. We propose an empirical parameterization of the organic-bearing particle concentration, with a dependence on water salinity and sea-surface temperature only. We also show that a very rich mixture of organic aerosols is emitted from the sea surface. Such data will certainly contribute to providing further insight into the influence of aerosols on cloud formation, and be used as input for the improved modeling of aerosols and their role in global climate processes.

summarizes the expedition route and specific events. Along this trajectory, dynamic, chemical, and biological characteristics of the ocean and the overlying atmosphere were continuously monitored using several instruments, as summarized in Table S2. Aerosols have been characterized by two complementary instruments operating in parallel. The two instruments are designed to sample air on the front side of the deck of the ship with a common access, shielded from wind by a cap.

Single particle fluorescence spectrometer
A single-particle fluorescence spectrometer (FPSF) was developed specifically for the present campaign in order to identify in real-time the aerosol particles present in the marine atmosphere. As sketched in Figure S1, the identification was achieved by measuring the scattering properties and the fluorescence spectrum of individual aerosol particles embedded in the atmospheric flow. In order to achieve this, the air was sampled at 60 L/min 5 m above sea level, concentrated by a 10x aerodynamic lens, and then focused by an optimally-designed sheath nozzle in a narrow stream of 500 µm diameter ( 1 ). This stream crosses infrared lasers for the scattering measurements, as well as a UV laser for exciting the fluorescence (2,3) that is detected by a multi-angle laser scattering module (4). This procedure also allows measuring simultaneously the optical size of each particle. Active "on the fly" modulation of the laser intensity is also applied so that saturation is prevented and particles whose sizes range from less than 1 µm to 60 µm can be analyzed. Scattering signals are then further used to trigger the pulsed UV laser (337 nm), which excites the fluorescence of the same individual particle. The fluorescence is retrieved by a dedicated reflective lens and analyzed with a 32-channel spectrometer from 360 nm to 650 nm. This spectrometer is based on a multi-anode photomultiplier (Hamamatsu H7260-03) and specifically-designed acquisition electronics. Figure S1. Principle of the single-particle aerosol spectrometer. The instrument allows measuring simultaneously multi-angle elastic scattering and the UV-Vis fluorescence spectra of individual aerosols The particles were sorted into clusters displaying similar fluorescence spectra, in a self-referenced procedure. A particle with a spectrum matching a previously-identified species is considered as a positive identification of the particle. Conversely, a particle matching no previously-identified cluster spectrum is considered as the prototype of a new cluster. The procedure was repeated after averaging the spectrum of each cluster over all its particles, in order to avoid overweighting the spectrum of the first particle within each cluster. The population of each cluster was then accumulated over hourly intervals. This self-referencing method avoids bias from a-priori selected reference spectra, to which experimental spectra are compared. This method was successfully applied with a similar aerosol fluorescence spectrometer developed at Yale (3), in the context of other environments, like urban pollution.
In parallel, we measured in the laboratory with the same instrument the fluorescence signatures of aerosols containing chemical compounds or organisms likely to be found in the ocean during the expedition. For this purpose, we vaporized them into a spray with a size distribution as close as possible to that measured during the campaign, and measured their fluorescence spectra in the GAP-SPFS operating in its standard mode. These spectra were averaged over several tens to several hundreds of individual particles per species. The inter-particle variability for each species was then used as a reference for the variability allowed in the cluster identification procedure.

Optical aerosol sizer
An optical aerosol sizer (Grimm 1.109) continuously sampled the air 5 m above sea level and measured the aerosol size distribution by optical scattering. It sorted the particles in 31 classes ranging from 250 nm to 32 µm. Note that this on-line measurement concerns the real aerosol size, as particles are not dried prior to, nor during, the measurement. The instrument was equipped with a nanoparticle sensor (Grimm Nanocheck 1.365) measuring the number concentration of particles in the range 25 -300 nm. This instrument samples 1.5 L/min through a tube that is 4 mm in diameter and ~1 m long. Measurement were recorded every 6 s, with a subsequent hourly averaging.

Conductivity, Temperature, Depth (CTD) vertical profiles
49 CTD profiles using an Idronaut Ocean Seven 316 Plus CTD, equipped with sensors for pressure, temperature, conductivity and oxygen, a Trilux sensor, which measures Chl-a, phycocyanine and phycoerythrin fluorescence and functions as a nepheolometer, and a Licor LI-193SA quantum sensor. This CTD was lost at sea (July 14th) and replaced by a Minos X, equiped with conductivity, temperature and pressure sensors (C•Xchange, T•Xchange, P•Xchange, respectively) for the remainder of the cruise.

Ferrybox
A Ferrybox continuously monitored the surface water by pumping water from approximately 1 m below sea surface at the rear of the ship. The water was analyzed in an Idronaut Ocean Seven 316 Plus, equipped with the same sensors as the CTD described above. The Chlorophyll a measurement was tested for consistency against satellite data (See Data validation below)

Datalogger
The ship was equipped with a driving assistance system and various sensors that were continuously logged. In addition to the technical parameters specific to the ship's navigation and performance (GPS location, heading, speed over ground and water, propulsion power), the recorded parameters included solar irradiance, air temperature, pressure, and wind speed and direction as well as sea temperature measured 1 m below the surface.

Weather station
In addition to its own independent measurements of atmospheric parameters, the ship was also equipped with a weather station (Batos, MeteoFrance), which continuously monitored wind speed, atmospheric pressure, air temperature and relative humidity, as well as the sea surface temperature. It also independently recorded the ship's trajectory, heading and speed.

Satellite data
We used the AQUA MODIS satellite 8-day composite Sea-Surface Temperature (SST) and chlorophyll a concentration data ( 5 , 6 , 7 , 8 ) to both request specific positioning of the ship during the expedition and also to cross-check the locally measured environmental parameters.

Data validation
All instruments were synchronized with a master clock for the entire duration of the campaign. The location of each measurement was determined according to this master clock, associated with the ship's trajectory data. After the validation and quality control specific for each instrument, data were averaged over hourly time intervals.
In a second step, we checked the consistency between the instruments whenever possible. This concerns in particular the atmospheric conditions (temperature, relative humidity, wind) from the ship data-logger and the Batos weather station; total aerosol concentrations from the GAP-SPFS and Grimm aerosol detectors, over the 1 µm -32 µm size range where both are sensitive (e.g., R = 0.69, p < 10 -6 for the 1-2.5 µm size range); and marine parameters (sea surface temperature, Chlorophyll a) between the Ferrybox and AQUA-MODIS data (R = 0.97, p < 10 -6 and R = 0.45, p < 10 -6 , respectively, see also Figure S2). However, as the measurement conditions are not identical, we did not use the data of Figure S2 as a calibration of the Ferrybox measurement of Chl a, that are provided as arbitrary units. The consistency with the Ferrybox with the first layers of the CTD profiles was also checked for SST (R = 0.999, p < 10 -6 ), and salinity (R = 0.73, p < 10 -6 ).  Figure S3. Dependence of the concentration of four particle size ranges on the atmospheric conditions (air temperature and wind speed, a -d) and oceanic conditions (salinity and sea surface temperature, e -h). Data represent hourly averages for each parameter during the expedition.  Table S3: Statistical comparison between sea surface properties and particle abundance and size. Pearson's correlation coefficient, p value and number of samples used for calculating the correlation are provided by hourly averages of the parameters collected by the Ferrybox (temperature, conductivity, and Chl a) and aerosols in the 1 -10 µm size range as measured by the optical aerosol sizer. Statistically significant positive correlations are highlighted in red, negative ones in blue. In order to verify the discrimination capability of the GAP-SPFS for some biological samples that may be present in the ocean, we performed laboratory tests using an atomizer and reference organisms. Figure S5 clearly demonstrates the capability of recognizing the characteristic broad spectrum from humic acid or the yellow peaked fluorescence of vitamin B2 (Riboflavin). The cyanobacterium Synechococcus and the planktonic diatoms Chaetoceros, exhibit, among others, the characteristic fluorescence of NAD(P)H in the blue and some Flavin related components in the yellow/red. It is interesting to notice that both the species and the nutriment condition (e.g. Iron rich or Iron poor) influence the fluorescence spectra. This was also previously observed for other bacteria (2)