Cloud processing is known to generate aerosol species such as sulfate and secondary organic aerosol, yet there is a scarcity of airborne data to
examine this issue. The NASA Aerosol Cloud meTeorology Interactions oVer the western ATlantic Experiment (ACTIVATE) was designed to build an
unprecedented dataset relevant to aerosol–cloud interactions with two coordinated aircraft over the northwestern Atlantic, with aerosol mass spectrometer data used from four deployments between 2020–2021 to contrast aerosol composition below, in (using a counterflow virtual impactor)
and above boundary layer clouds. Consistent features in all time periods of the deployments (January–March, May–June, August–September) include
the mass fraction of organics and relative amount of oxygenated organics (
The nature of aerosol–cloud interactions over the northwestern Atlantic Ocean is uncertain even though the region has been the target of decades of atmospheric research (Sorooshian et al., 2020). These interactions include a subset of aerosol particles called cloud condensation nuclei (CCN) that activate into cloud droplets, which subsequently undergo aqueous processing to transform into a particle after evaporation varying in size and composition relative to the original CCN. An aspect of these steps that is poorly characterized is the composition of the droplet residuals in cloud relative to particles below and above clouds, which requires airborne measurements. The NASA Aerosol Cloud meTeorology Interactions oVer the western ATlantic Experiment (ACTIVATE) was designed to collect in situ and remote sensing data in and around clouds during different seasons in a region with a wide range of weather conditions (Painemal et al., 2021) and air mass sources (Corral et al., 2021), qualifying as a suitable dataset to examine this very issue.
The annual cycle of aerosol and cloud drop number concentrations (
The goal of this study is to compare aerosol mass spectrometer data over the northwestern Atlantic below, in, and above clouds for different times of the year (January–March, May–June, August–September). Case studies of flights during cold air outbreaks probe deeper to better understand the nature of aerosol and droplet residual particle composition during these events with stronger aerosol–cloud interactions as compared to other times of the year (Dadashazar et al., 2021a; Painemal et al., 2021). The results have implications for aerosol–cloud interactions as droplet residual composition is shown here to deviate from that of aerosol out of cloud.
We use airborne in situ data collected aboard the HU-25 Falcon from deployments 1 (14 February–12 March 2020), 2 (13 August–30 September 2020),
3 (27 January–2 April 2021), and 4 (13 May–30 June 2021) of the ACTIVATE mission. Data necessary for this study were only available for two flights
in deployment 3 (29 January and 3 February) owing to an aircraft maintenance issue reducing the size of the available payload. ACTIVATE employs a dual
aircraft approach with the Falcon acquiring in situ data for trace gases, aerosol particles, and clouds in the MBL while a King Air flies overhead at
The central dataset relevant to aerosol composition in this study comes from the aerodyne high-resolution time-of-flight aerosol mass spectrometer
(AMS) (DeCarlo et al., 2008). The instrument measures submicrometer non-refractory aerosol composition in 1
Note that while cloud water samples were also chemically characterized, those data are outside the scope of this work as the partial speciation of
organics in the cloud water samples makes it hard to compare to AMS total organics. Furthermore, particle-into-liquid sampler (PILS) data are not used
owing to lengthier time resolution (
We obtained 5
As this study is mainly focused on organics and sulfate, concentration-weighted trajectory (CWT) maps were generated using HYSPLIT back-trajectories in conjunction with speciated AMS data (Figs. S1 and S2 in the Supplement) to show the predominant sources for each of these two aerosol components (e.g., Hsu et al., 2003). As demonstrated by past works for other regions (e.g., Dadashazar et al., 2019), the method assigns a weighted concentration to grid cells based on mean concentrations passing through each grid cell from all the considered trajectories. CWT profile maps are produced using the GIS-based software called TrajStat (Wang et al., 2009).
Spatial map of cloud-free AMS data for organics and sulfate collected during deployments 1–4 of ACTIVATE spanning from February 2020 to June 2021. Non-CAO and CAO represent non-cold air outbreak and cold air outbreak days, respectively, between January and March. Spatial boxes labeled 1–3 in
We use both total and speciated (sulfate and organic) aerosol optical depth (AOD) at 550
Cloud droplet number concentrations (
We determine whether flights occurred during cold air outbreaks (CAOs) leveraging methods in recent ACTIVATE studies (Seethala et al., 2021; Corral et al., 2022). Briefly, Visible Infrared Imaging Radiometer Suite (VIIRS) imagery (NASA Worldview) is used to visually identify cloud streets that are characteristic of CAOs. Flight notes and weather forecast slides were used as additional confirmation, followed by data from dropsondes released from the King Air following the method described in Papritz et al. (2015).
A motivation of this study is the opposite annual pattern of
Vertically-resolved cloud-free AMS data for the different time periods of ACTIVATE deployments and boxes defined in Fig. 2a. Shown are (left to right) organic and sulfate concentrations, organic and sulfate mass fraction, and the ratio of
Relative to all AMS species, sulfate and organics are the dominant aerosol components by mass with combined mass fractions being near 75 % usually
regardless of season or location relative to clouds (Tables S1 and S2 in the Supplement; spatial maps
in Fig. 2); this is consistent with their predictive capability for
In contrast to organics, sulfate exhibits more spatially homogenous concentrations over the northwestern Atlantic (Fig. 2) owing largely to ocean-emitted dimethylsulfide that undergoes gas and in-cloud oxidation such as what was shown for the eastern North Atlantic (Ovadnevaite et al., 2014). This is supported by how sulfate's seasonal CWT maps (Fig. S2) differ from those of organics with comparable concentrations widespread over the northwestern Atlantic relative to the continent. The August–September CWT map for sulfate reveals more high concentration areas (note the different color bar scale for August–September in Fig. S2) over the continent with concentrations exceeding those over most of the ocean; this is presumably due to more secondary formation stemming from local sulfur dioxide emissions over the eastern US (Yang et al., 2018) aided in part by higher temperatures and humidity (Corral et al., 2021) that co-vary with other conditions favorable for sulfate production such as stagnation and certain air flow patterns (Tai et al., 2010). Figure 3 demonstrates that neither sulfate nor organics exhibit a clear reduction with altitude pointing towards a potential source aloft that might include long-range transport and/or secondary production.
Although based on only two consecutive days of flight data, results from Leaitch et al. (2010) are relevant in that they sampled below, in, and above boundary layer clouds over the northwestern Atlantic. On the first day with more marine influence, sulfate was more abundant than organics in fine particles below cloud. In contrast, the second day had more continental influence with organic levels exceeding those of sulfate below cloud, which was often the case during ACTIVATE (Table S1). They concluded with a parcel model that the impact of anthropogenic carbonaceous components on the cloud albedo effect can exceed that of anthropogenic sulfate, which motivates attention to the droplet residual composition discussed next.
Seasonal comparison of AMS mass fractions, including the relative contribution of
A striking result in all seasons is that organic mass fraction was higher downstream of the CVI in droplet residual particles in contrast to adjacent
BCB and ACT legs in cloud ensembles (Fig. 4 and Table S1). To compensate, sulfate mass fractions decreased in droplet residuals. Furthermore,
The organic mass fraction and
Scatterplot of the difference in organic mass fraction in cloud legs with CVI data and below cloud base (BCB) legs for an individual cloud ensemble relative to the analogous difference for sulfate mass fraction between the same pair of legs. Markers are colored by the analogous difference in nitrate mass fraction. Panels represent different seasons with winter deployments (January–March) separated into CAO and non-CAO days.
We next examine scatterplots of organic mass fraction (i.e., organic mass divided by total AMS mass) differences between each cloud leg with CVI–AMS
data and its closest BCB leg in the same cloud ensemble versus analogous sulfate mass fraction differences for the same pair of legs (Fig. 5). Aqueous
processing to preferentially increase one of the two species relative to the other would presumably translate into a positive value on the more
preferred species' axis; in other words, if there was more organic aerosol formation in clouds via aqueous processing relative to sulfate, it would
register as a positive (negative) value on the
A comparison of
Comparison of
The CVI droplet residuals are more oxidized because of some combination of aqueous processing effects to yield more oxidized organic species, or
because CCN with higher
Scatterplot of the difference in
A discussion on possible contributing factors (other than aqueous processing) to the different chemical signature in CVI samples relative to adjacent cloud-free areas is warranted. First, we note that 23 % of BCB/CVI pairs of data points (25 out of 110) exhibited higher organic mass fraction in the BCB leg relative to droplet residuals (Fig. 8). This number increases to 26 % when considering if either the BCB or ACT organic mass fraction was higher than the corresponding CVI data in cloud for an ensemble. Clearly the cases where a higher organic mass fraction was observed out of cloud seems to be most prevalent below cloud suggesting that location is where a cloud processing signature can be more reliably observed. These 26 % of the cases studied demonstrate that the null case exists without an organic enhancement downstream of the CVI, reducing concerns over instrument and sampling artifacts.
Scatterplot of organic mass fraction in droplet residuals (downstream CVI in cloud) and in aerosol sampled during the closest below cloud base (BCB) leg from ACTIVATE deployments 1–4. A total of 25 points out of a total of 110 (23 %) were below the
In terms of the contamination due to the inlet's material of construction, the CVI inlet was designed with both stainless steel and aluminum yielding
negligible organic contamination (Shingler et al., 2012). A way to test this is to conduct CVI sampling in cloud-free conditions. Figure S3 in the
Supplement shows a representative time series of AMS data during a flight (research flight 10 on 28 February 2020) with numerous cloud passes and
periods when there was still sampling downstream of the CVI inlet outside of cloud. During those three key periods shown out of cloud with CVI
sampling, sulfate and organic levels exhibit concentrations close to zero and with concentrations considerably lower than CVI data in cloud. Compared
to sulfate, there is more variability in organic levels downstream of the CVI regardless of whether sampling was in or out of cloud or even whether
sampling was done using the isokinetic inlet out of cloud. The data reveal that at small timescales there is variability in the
The heated counterflow in the CVI reduces positive artifacts from volatile gaseous species partitioning into sampled droplets such as with volatile organic compounds (VOCs) to form organics or with nitric acid to form nitrate (Prabhakar et al., 2014); in contrast, the heated counterflow would presumably evaporate some fraction of the existing nitrate and organics in the CCN that activated into droplets unlike sulfate which is not volatile. Thus, the heated inlet would tend to favor sulfate in the cloud droplet residuals and could not explain the enhanced organic residual observations here.
Inlets including the CVI can be prone to droplet shatter such as with large drizzle drops (
It is also noteworthy that there can be considerable variability in AMS composition along level legs (BCB, in cloud, ACT) pointing to how a signature
of cloud processing out of cloud can be reduced when averaging data. Figure S3 demonstrates variability along individual legs that is not consistent
with the
The previous discussion does not provide support for any form of artifact or contamination explaining why 74 % of the CVI data points exhibited higher organic mass fractions than both the BCB or ACT legs. One could argue that the chemical signature of cloud processing should be evident out of cloud somewhere as ultimately the droplet residual particles will evaporate outside of cloud and return to the aerosol phase. As will be discussed in Sect. 4 though, there is a body of literature pointing to droplet residuals having the strongest signature of cloud processing rather than below or above cloud. Although difficult to prove with this dataset, a plausible explanation is that the processed aerosol dilutes into the MBL at a timescale that is much faster than the production/evaporation cycle.
Summary of AMS composition in adjacent BCB, cloud, and ACT legs during back-to-back flights (research flights 17 and 18) in CAO conditions on 8 March 2020. Shown in the bar charts are the mass fractions of AMS components in addition to either total AMS mass (for ACT and BCB legs; such data are not robust for CVI legs due to how the CVI operates) or longitude on the right
Summary of AMS composition in adjacent BCB, cloud, and ACT legs during back-to-back flights (research flights 5 and 6) in CAO conditions on 22 February 2020. Shown in the bar charts are the mass fractions of AMS components in addition to either total AMS mass (for ACT and BCB legs; such data are not robust for CVI legs due to how the CVI operates) or longitude on the right
Summary of AMS composition in adjacent BCB, cloud, and ACT legs during back-to-back flights (research flights 10 and 11) in cold air outbreak conditions on 28 February 2020. Shown in the bar charts are the mass fractions of AMS components in addition to either total AMS mass (for ACT and BCB legs; such data are not robust for CVI legs due to how the CVI operates) or longitude on the right
Owing to interest in the winter season having the strongest aerosol–cloud interactions (Dadashazar et al., 2021a; Painemal et al., 2021), here we
examine six case study research flights (RFs) during CAOs to understand the compositional characteristics below, inside, and above clouds. We focus
more on the representative day of 8 March 2020 (Fig. 9), which included two consecutive flights (RFs 17 and 18) based out of Hampton, Virginia
profiling aerosol and cloud properties in CAO conditions. These two flights were investigated in past work showing enhanced new particle formation in
ACT legs (Corral et al., 2022) and that entrainment of free tropospheric air dilutes MBL CCN concentrations (Tornow et al., 2022). The other four flights (Fig. 10: RFs 5–6 on 22 February 2020; Fig. 11: RFs 10–11 on 28 February 2020) exhibited
the same general results as those shown for 8 March with higher organic mass fractions and
Figure 9 shows the AMS composition profile on the out-and-back flights on 8 March, which involved flying out to a point and repeating the same path
back to the airfield. Stacked on top of each other in Fig. 9 are the corresponding legs within individual cloud ensembles including (from top to
bottom) ACT, either BCT or ACB legs with CVI data, and BCB. RF17 in the morning comprised 13 different cloud legs with corresponding BCB and ACT
legs. The BCB and ACT mass fraction profiles were similar with sulfate being most abundant (mass fractions: 0.34–0.65) followed closely by organics
(mass fractions: 0.15–0.42). The
Our results represent unique atmospheric data that are scarce in the literature owing to the difficulty of obtaining aerosol chemical data below, in,
and above cloud in close spatiotemporal proximity across many flights in different times of the year. Figure 1 provides implications of the results in
terms of differences with MERRA-2 speciated AOD. Although we cannot unambiguously prove it with the dataset, the results suggest that the most likely
explanation for organic and
Coggon et al. (2012) showed increased AMS
While a measurement of hygroscopicity of the droplet residuals was not available, we instead examine aerosol hygroscopicity from BCB legs as that is the area out of cloud most commonly exhibiting higher organic mass fractions relative to in cloud. Even if the signature out of cloud is not as clear as one would expect presumably owing to dilution effects, still the influence of cloud processing on organics inevitably should exist to some extent making the subsequent discussion valuable. Having more organics relative to sulfate may reduce hygroscopicity at high RHs (e.g., Hersey et al., 2009), but a compensating factor could be that the organics are more oxygenated, which would increase the hygroscopicity of the organic fraction itself.
Figure 12 shows an inverse relationship between
Relationship between
A large airborne dataset collected over the northwestern Atlantic as part of the NASA ACTIVATE mission reveals a distinctly different chemical signature in cloud droplet residuals (lower sulfate mass fraction, higher organic mass fraction, and higher
The results of this study motivate increased attention to both in-cloud formation of oxygenated organics and the composition of particles activating into droplets over the northwestern Atlantic. This work has implications for aerosol–cloud interactions in this region as datasets often relied on in the absence of airborne data such as reanalysis data suggest a different story where sulfate is more enhanced than organics year-round (in contrast to the airborne data) (e.g., Braun et al., 2021). The high relative abundance of organics needs more attention, especially in light of the increasing relative amount of species in aerosol particles other than sulfate due to regulatory activities over the US (Hand et al., 2012).
ACTIVATE airborne data:
The supplement related to this article is available online at:
HD conducted the analysis. AS and HD prepared the paper. All authors contributed by providing input and/or participating in airborne data collection.
At least one of the (co-)authors is a member of the editorial board of
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This article is part of the special issue “Marine aerosols, trace gases, and clouds over the North Atlantic (ACP/AMT inter-journal SI)”. It is not associated with a conference.
The work was funded by NASA grant 80NSSC19K0442 in support of ACTIVATE, a NASA Earth Venture Suborbital-3 (EVS-3) investigation funded by NASA's Earth Science Division and managed through the Earth System Science Pathfinder Program Office. We acknowledge use of imagery from the NASA Worldview application (
This research has been supported by the National Aeronautics and Space Administration (grant no. 80NSSC19K0442), the Helmholtz Association (grant no. HGF W2W3-060), and the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) – TRR 301 – Project-ID 428312742.
This paper was edited by Markus Petters and reviewed by two anonymous referees.