Giant Sea Spray Particles Found to Accelerate Warm Rain in Marine Clouds

High above the western North Atlantic, the ocean is quietly seeding the clouds above it with particles large enough to change when rain begins. A new analysis of aircraft measurements from the NASA ACTIVATE campaign, published in Atmospheric Chemistry and Physics, shows that coarse-mode sea spray aerosols — salt particles with dry diameters exceeding roughly 2 micrometers — can measurably accelerate the formation of warm rain in marine low clouds. These particles, known as giant cloud condensation nuclei, or GCCN, act as the starting points for the oversized collector droplets that clouds need to initiate collision-coalescence, the process by which droplets merge and grow into drizzle. The study, led by Katherine Mifsud of the University of Washington together with colleagues at NASA Langley Research Center, the University of Arizona, and the German Aerospace Center, combines an unusually rich observational record with a state-of-the-art super droplet model to argue that even modest amounts of giant sea salt can tip lightly precipitating clouds toward earlier and heavier drizzle.

The stakes of this finding extend well beyond cloud physics trivia. Marine stratocumulus clouds cover roughly 20 percent of the planet’s oceans and play a dominant role in modulating Earth’s radiative energy balance, reflecting a large portion of incoming solar radiation while contributing only minimally to outgoing longwave radiation. How these clouds respond to aerosol perturbations is one of the largest and most stubborn uncertainties in projections of future warming. The classic Twomey effect describes how added aerosols increase droplet concentration, shrink droplet size for a fixed amount of condensate, and brighten clouds. But changes in droplet size also drive adjustments to cloud lifetime, thickness, and precipitation, and these adjustments remain poorly quantified in Earth system models because they unfold on sub-kilometer scales that models cannot resolve. Precipitation processes, in particular, are a persistent source of uncertainty, which is precisely where the new study aims to make a difference.

The physical mechanism at the heart of the work has a solid theoretical pedigree but has never been fully incorporated into the bulk microphysical parameterizations used in most climate models. Droplets formed on GCCN are highly concentrated salt solutions compared with the majority of cloud droplets, which grow on much smaller accumulation-mode particles with typical dry diameters of about 0.1 micrometers. This concentration difference accelerates, via the Raoult effect, the condensational growth rates of GCCN-containing droplets, allowing them to reach radii of 15 to 20 micrometers — the size at which they can begin sweeping up smaller droplets through collision-coalescence. Because these salty droplets resist evaporation, the larger ones can even continue growing by condensation as they descend through weakly subsaturated downdrafts, conditions in which ordinary droplets typically lose mass and may evaporate entirely. The modeling challenge is that a droplet’s condensational growth rate depends not only on its size but also on the mass of salt it contains, meaning GCCN-containing droplets must be treated as a distinct category of hydrometeor — a computational burden that super droplet methods can now shoulder.

To ground the modeling in reality, the team drew on in-situ observations collected during two deployments spanning January to June 2022, when the ACTIVATE campaign flew a twin-aircraft design out of NASA Langley Research Center in Virginia. The low-flying HU-25 Falcon, which sampled altitudes between 0.15 and 3 kilometers, carried a suite of externally mounted probes including the Cloud Aerosol Spectrometer, the Cloud Droplet Probe, the two-dimensional Stereo Probe, and the Diode Laser Hygrometer. In total, 75 joint flights on 43 days logged upwards of 120 flight hours during the 2022 deployment, making it the most comprehensive year of the campaign. The researchers focused on below-cloud-base flight legs, applying strict filtering criteria — a liquid water content threshold, a cap on precipitation-sized particles, and a relative humidity ceiling of 95 percent — to ensure that the measured hydrated size distributions characterized haze particles in subsaturated air, free of cloud and drizzle contamination. After filtering, 460 below-cloud-base legs remained, each roughly three minutes long, with a mean altitude of 701 meters.

Converting the measured ambient distributions to dry salt size distributions required accounting for hygroscopic growth: ambient particle diameters were divided by a growth factor derived from each leg’s mean relative humidity, following a formulation based on laboratory measurements of sea salt. The resulting average dry size distributions follow an approximately exponential decay with increasing diameter, with a tail extending to about 20 to 25 micrometers. The mean total number concentration of coarse-mode particles was 0.56 per cubic centimeter for the CAS and 0.18 per cubic centimeter for the CDP, while mean dry salt masses were 10.3 and 12.2 micrograms per cubic meter respectively. The team then examined how these quantities depend on near-surface wind speed, standardizing all aircraft wind measurements to 10 meters above sea level using a logarithmic wind profile correction. The mean corrected wind speed across all valid legs was 5.31 meters per second with a standard deviation of 2.65 meters per second.

Binned into six wind speed ranges spanning 0 to 16 meters per second, both the number and mass concentrations of coarse-mode sea salt increased with wind speed, consistent with the established picture that stronger winds enhance bubble bursting and sea spray production at the ocean surface. The number concentration–wind speed fits yielded coefficients of determination of 0.84 for the CAS and 0.64 for the CDP, while the mass–wind speed relationships were weaker, at 0.61 and 0.54. Yet substantial variability remained within each wind speed bin, and the observed dependence was notably weaker than would be expected from sea spray aerosol flux parameterizations, which typically show a much steeper increase in particle production with wind speed. The comparison with earlier work was encouraging: the slopes and wind speed ranges overlapped with results from the RICO campaign over the trade winds, and the observed mean salt masses fell within the regime reported by Woodcock in 1953. Still, the authors conclude that wind speed alone does not fully explain GCCN variability, and no clear seasonal trend emerged across the sampled months — properties appeared driven more by individual flight conditions than by a coherent seasonal cycle.

The modeling half of the study used a one-dimensional kinematic super droplet model, based on the Shima framework, in which a column of air is lifted, haze particles grow by condensation, a fraction activates into cloud droplets, and collision-coalescence and rainout proceed over a one-hour simulation — a typical cloud lifetime. The simulated cloud was 400 meters thick, with a peak updraft speed of 1 meter per second that linearly decayed to zero by 800 seconds, producing a liquid water path of about 385 grams per square meter, representative of the thicker parts of marine boundary layer clouds likely to precipitate. The background submicron aerosol concentration was set to 50 per cubic centimeter, typical of relatively clean marine conditions, with sensitivity tests spanning 10 to 100 per cubic centimeter. Each case used an ensemble of 16 runs with 20,200 super droplets, and all 460 observed flight-leg size distributions served as model inputs for the GCCN population.

The simulations revealed a clear and consistent enhancement of accumulated rainfall with increasing GCCN mass and number concentration across all model ensembles. Crucially, even relatively small GCCN salt masses of around 1 microgram per cubic meter could accelerate the onset of drizzle in lightly precipitating clouds. The background aerosol environment set the effectiveness of the giant particles: under low aerosol conditions of 10 per cubic centimeter, clouds produced the most rainfall for a given GCCN loading, because fewer competing smaller droplets made collision-coalescence more efficient, while high-aerosol ensembles distributed available water across many droplets and slowed rainfall. The sensitivity of rain to GCCN was strongest, in a relative sense, at high background aerosol concentrations. Mass concentration proved a tighter predictor of rainfall than number concentration, particularly for thin clouds with a liquid water path of 100 grams per square meter, where rain remained vanishingly small until GCCN loadings rose significantly — evidence that GCCN alone cannot generate rainfall when liquid water is limited. Microscale turbulence, tested at two dissipation rates, increased rain rates across all ensembles but remained secondary to the effects of background aerosol and cloud water. Under comparable conditions, the model’s rainfall magnitudes agreed within roughly a factor of three with an independent parcel-model study by Jensen and Lee.

The observational analysis provided an independent check. Using one-second in-cloud data, the team computed rainwater content from the 2D-S probe for drops larger than 50 micrometers, cloud water content from the CAS, and the ratio of rainwater to total condensed water as a proxy for precipitation efficiency. Flights were split into high and low GCCN categories based on flight-mean values, and a bootstrap resampling approach with 10,000 realizations per bin assessed statistical significance within bins of liquid water content and droplet number concentration. Higher GCCN conditions were associated with enhanced rainwater fractions in the vast majority of bins for both mass and number metrics, with the enhancement especially robust in the central, heavily populated bins at moderate liquid water content and droplet concentration, where bootstrap realizations exceeded unity nearly 100 percent of the time. In several densely sampled bins, the rainwater fraction was enhanced by approximately a factor of two or more. The observations could not clearly distinguish whether mass or number concentration was the better predictor, partly because the two flight-mean metrics covary, but they firmly supported the model’s central conclusion.

Together, the results point toward a practical path forward: because the influence of GCCN appears well captured by total dry salt mass, future parameterizations could represent these effects without tracking full size distributions. The authors caution that the ACTIVATE aircraft carried no radar, so the in-situ rainwater measurements cannot capture precipitation reaching the surface, and mass estimates remain sensitive to the rare, largest particles that dominate mass but are hardest to sample. The treatment of turbulence in the collision kernel also introduces uncertainty, though the authors express confidence in the robustness of the non-turbulent cases. Even so, the consistency between observations and simulations delivers a significant message for climate science: coarse-mode sea spray can enhance precipitation efficiency in marine low clouds at concentrations far lower than previously appreciated, and representing this effect accurately may be essential for reducing the persistent uncertainty in how clouds and precipitation respond to aerosols — a key lever on the pace of future warming.

Subject of Research: Coarse-mode sea spray aerosols enhance warm rain formation in marine low clouds

Article Title: Coarse-mode sea spray aerosols enhance warm rain formation in marine low clouds

Article References: Mifsud, K., Barr, J., Wood, R., Choi, Y., Crosbie, E., Diskin, G. S., DiGangi, J. P., Moore, R. H., Robinson, C., Shook, M. A., Sorooshian, A., Voigt, C., & Ziemba, L. (2026). Coarse-mode sea spray aerosols enhance warm rain formation in marine low clouds. Atmospheric Chemistry and Physics, 26(20), 14881-14903. https://doi.org/10.5194/acp-26-14881-2026

Image Credits: AI Generated

DOI: 10.5194/acp-26-14881-2026

Keywords: sea spray aerosols, marine stratocumulus, warm rain, giant cloud condensation nuclei, ACTIVATE campaign, aerosol-cloud interactions, drizzle, climate modeling, Coarse-mode, spray, aerosols, enhance