Planar intrusions from a narrow, jet-like source in a linearly stratified ambient

In the deep ocean, in the atmosphere above us, and even within the earth's crust, fluids rarely sit still. They move, mix, and push against one another, often carrying heat, salt, or ash from one place to another. When a fluid that is lighter or heavier than its surroundings is released, it does not simply sink or rise in a straight line; it seeks a level where its density matches the environment. At this level of neutral buoyancy, the fluid spreads out horizontally, forming a distinct layer that intrudes into the surrounding medium. These intrusions are vital to understanding how energy and matter travel through our planet, from the way volcanic ash clouds settle to how pollution disperses in the sea. For decades, scientists have studied how these layers form when released from a large, wide opening, where the flow behaves in a predictable, steady manner. However, nature often releases fluids from narrow cracks, pipes, or vents, and the rules governing these narrow sources have remained less clear, leaving a gap in our understanding of how these critical flows begin and evolve.

To fill this gap, researchers at Monash University in Australia turned to powerful computer simulations to watch, in extreme detail, what happens when a narrow jet of fluid is injected into a layered, stratified environment. They modeled a scenario where a continuous stream of fluid, perfectly mixed and ready to spread, is forced out of a thin slot into a surrounding fluid that gets steadily denser with depth. By running these simulations on supercomputers, they could track the fluid's journey from the very first moment of release through its chaotic middle stages and into its final, stable form. The team performed these experiments in both two dimensions, which allowed them to test many different conditions, and in three dimensions, which revealed the true, complex behavior of the flow as it would occur in the real world.

The story of this intrusion unfolds in three distinct chapters. In the beginning, the fluid shoots out like a jet, but it immediately organizes itself into a leading structure: a pair of spinning vortices, one rotating clockwise and the other counter-clockwise, locked together like a single unit. This vortex dipole acts as the front of the intrusion, pushing forward at a steady speed. Trailing behind this leading edge is a thin stream of fluid that ripples with waves. In the early moments, these waves are symmetrical, rising and falling in unison on both the top and bottom of the stream. As the flow continues, the environment begins to change. The stratification of the surrounding fluid, which acts like a series of invisible horizontal layers, squeezes the vortex dipole, preventing it from growing too tall, but it does not stop the dipole from moving forward. In fact, the more strongly layered the environment is, the faster this leading dipole travels, as the vertical confinement forces the energy to push outward horizontally.

The second chapter begins when the trailing stream of fluid becomes unstable. The symmetrical waves that once rose and fell together begin to break their symmetry. One side of the stream starts to flap up while the other flaps down, a motion that grows rapidly and spreads both backward toward the source and forward toward the leading dipole. This is a classic instability, driven by the sheer speed difference between the moving fluid and the still fluid around it, rather than by the density layers themselves. The researchers found that this chaotic flapping happens with a specific rhythm and speed that matches what is predicted for simple, unlayered jets, suggesting that the initial instability is a local event determined by the flow's own speed and width. As this instability grows, it tears the smooth stream apart, mixing the fluid and creating a turbulent wake that eventually engulfs the entire intrusion.

The final chapter reveals the true nature of the intrusion once the chaos settles. In the two-dimensional simulations, the flow eventually smooths out into a wedge shape, thick near the source and tapering to a point at the front. However, the three-dimensional simulations told a different and more revealing story about the region near the source. In the real, three-dimensional world, the area right next to the injection point does not stay thin and uniform. Instead, it becomes a zone of intense, active mixing. The fluid here expands and swirls, pulling in a significant amount of the surrounding environment—up to one hundred percent more fluid than was originally injected. This near-source region acts as a powerful engine, entraining the ambient fluid and expanding the flow's width before it even reaches the main body of the intrusion.

Once this expanding, mixing zone is accounted for, the rest of the intrusion behaves in a surprisingly simple way. The far-field part of the flow, the long wedge stretching out ahead, turns out to be indistinguishable from intrusions that started from a wide source. The narrowness of the original pipe or crack does not dictate the shape of the final intrusion; rather, the narrow source acts as a feeder that creates a specific amount of effective flow through its mixing zone. The researchers found that if they measured this effective flow rate and used it to scale the data, the narrow-source intrusions collapsed perfectly onto the same mathematical description used for wide-source intrusions. This means that the complex, turbulent dance happening right at the source is what matters most; once the fluid leaves that zone, it forgets the narrowness of its origin and follows the universal rules of buoyancy-driven flow.

The study also clarified how the intrusion interacts with the surrounding environment. As the leading dipole moves, it pushes the layers of the surrounding fluid up and down, creating internal waves. Under conditions of weak ambient stratification, these waves can propagate faster than the intrusion front, and their interaction with the tip can compress it from behind, producing a thin, flattened tip and a pulsing modulation in the front's propagation. However, measurements from previous experimental studies suggested that despite these interactions, the internal waves had only a limited effect on the overall thickness profiles and propagation speed of the intrusion.

By combining high-resolution simulations with careful analysis, the researchers have provided a complete picture of how narrow intrusions evolve. They showed that the flow is not a single, uniform process but a sequence of distinct phases, each with its own dominant physics. The initial momentum-driven phase gives way to a shear-driven instability, which finally relaxes into a buoyancy-controlled state. Crucially, they demonstrated that the three-dimensional nature of the flow near the source is essential for understanding the whole system. Without accounting for the extra fluid pulled in by the expanding jet near the inlet, the behavior of the intrusion cannot be accurately predicted. This work bridges the gap between the simple models of wide sources and the complex reality of narrow vents, offering a unified framework that explains how fluid layers form and spread in our oceans and atmosphere. The findings suggest that to predict where these intrusions will go and how much they will mix, one must first understand the turbulent, expanding zone right at the source, where the fluid learns to become a wide, stable layer.