Sea spray droplets torn from wave crests during storms carry a film of contaminants and biofilms that changes how they break apart, and this project will build that chemistry into the models used to predict hurricane intensity. Current storm models cannot agree on how much spray is produced at high wind speeds, partly because they ignore the surfactants—soap-like molecules—that coat real seawater. These surfactants alter how droplets form and shatter, a process called bag-breakup that generates thousands of tiny fragments. Without accounting for this, forecasts of storm strength remain unreliable. If the team succeeds, weather agencies will have a sea spray generation function that includes surfactant effects. That should improve predictions of how fast a storm will intensify, giving port authorities and coastal councils more accurate warnings. The data will be hosted openly, designed for future use as training material for machine-learning storm models. This is fundamental fluid dynamics, not an immediate operational tool. But the same kind of small-scale physics—how a droplet bursts, how a film ruptures—has historically been the missing piece in larger models, from combustion engines to climate simulations. Better understanding of bag-breakup could eventually feed into any system where droplets and wind interact.
View original technical description
The intensities of severe storms at sea are difficult to predict. This is partly because sea-surface heat transfer and drag coefficients vary with increasing wind speed in ways not clearly understood. Spume – sea spray droplets torn directly by wind from wave crests - could plausibly cause these variations at high winds (>18m/s) despite its microscale origin. The total production of spray is estimated by sea spray generation functions (SSGFs), but existing SSGFs cannot agree on spume production rates. SSGFs also do not account for the physicochemical properties of the spray itself: seawater is abundant in biofilms and contaminants, which can be modelled as surfactants. Underpinning the challenge of modelling storms is their highly multiscale nature, with physical influences ranging from regional scales (105m) to droplets at the ocean surface (10-5m). Indeed, literature has shown that quantifying the dynamics of even the smallest droplets produced at the sea-surface to estimate SSGFs or other air-sea fluxes can explain much variability of existing data. This project will investigate the small-scale physics of surfactant-laden droplets and spume-producing waves, and ultimately generate an SSGF that accounts for surfactants in the seawater. Spume is known to be generated via a phenomenon called bag-breakup, but this is generally hard to investigate because it generates many tiny fragments in a process not fully understood. Surfactants further complicate these dynamics. In this project, state-of-the-art experimental and numerical approaches will be used to understand bag-breakup of surfactant-laden drops. First, the breakup of individual drops will be investigated, examining in detail the rupture of thin films and disintegration of larger, droplet rims. Second, the statistical occurrence of bag-breakup events on the surface of propagating, wind-driven waves will be measured in the laboratory and numerically. And third, the data from these droplet- and wave-scales will be combined with the theory of sea states to derive an SSGF for surfactant-laden droplets. The main impact of this work, delivered through scientific publications and conference focus sessions, will be to improve modelling of ocean storm intensity, thereby improving early-warning systems, ultimately saving lives. To achieve indirect impacts of maximising stakeholder access to scientific data, the data will also be hosted on an open platform freely available for use by others with particular attention towards future utility as training data for machine algorithms; these remain an avenue of future research for the growing machine-learning skills of the investigating team. Beneficiaries of this work will include the scientific community; weather modelling agencies; port authorities and coastal town councils; and ultimately the general public. This project will be in full alignment with several key NERC approaches on digital strategy aimed at maximising its impact. Firstly, cutting-edge numerical simulation on the UKRI ARCHER2 supercomputer is a core part of this project, maximising use of Computational Resources for Environmental Science. Secondly, it is aimed at Enhancing Data Services through digital hosting of data at the University of Oxford, as described above. And thirdly, the project ambition is to support Pioneering Computational Science not just through the high-fidelity numerical simulations above, but also in exploring ways to present data in ways most accessible to machine learning. In this connection, the investigating team’s internal aim is to Develop People and Skills locally in the Department, in data science to deliver on the project ambition.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know