Part III: The GLV Building Blocks as a Philosophical Framework. Layer 5 Cosmic Noise, Bursts, and Fluctuations

In the first four layers I treated the universe as if the lensing cosmic web were a kind of quiet background. The winding route of light, the direction sensitive distortion, the scale dependent amplification, and the build up through time all describe a web that does grow, but that changes fairly gradually on cosmic timescales. Layer 5 breaks with that picture. Here the dynamics come forward, the fact that the web sometimes does not grow gently, but jolts, grinds, and explodes.

Gravity is not only the slow architect of the universe, it is also the engine behind violence. Clusters falling into each other, galaxies colliding head on, halos merging and making one another tremble. In all those processes the mass distribution does not change millimetre by millimetre, but in jumps. The gravitational well deepens or shifts, substructures pull through one another, and the potential field around such a system is temporarily anything but quiet. Layer 5 is about what light paths do when they travel exactly through such a restless phase.

You can compare it to driving through a city under construction. The layers so far described the traffic jam, the average congestion, the fixed roadworks. Layer 5 is about the moments when a crane suddenly swings into the street, a row of trucks blocks an intersection, or an entire line of lanes is temporarily closed. If your route happens to pass such an incident, your travel time is not only determined by the average congestion, but by a temporary spike. If you add up many such journeys over a long time, a kind of background of extra delay and extra variation appears, on top of the average pattern.

Something similar happens in the gravitational field of the universe. During the formation of large clusters, during the merging of halos, and in other violent events, the lens strength along certain lines of sight is temporarily stronger or differently distributed than in the quiet periods between them. Light that travels past such a region at exactly that moment receives an extra tug. Those extra tugs are the gravitational bursts Layer 5 refers to. Not bursts of light, but bursts in the lensing action of spacetime.

What matters is that these bursts do not run neatly in sync. A cluster at distance A may be in a merger phase right now, while another cluster at distance B has already settled down. Across the entire visible three sphere there are always regions somewhere where such a gravity storm is raging. The sum of all those independently occurring peaks forms what I call the gravitational burst background, a diffuse noise layer of temporary extra lens strength, which lies on top of the slow, average lens stack of Layers 1 through 4.

What do you notice of this concretely in observations. A first effect is extra fine structure in lens maps. In weak lensing we reconstruct a map of the foreground mass from the statistical distortion of many distant galaxies. In a purely calmly growing web, that map would be relatively smooth, with gradually rising structures and broad peaks. In practice we often see a little more graininess and small finger like protrusions than you would expect from such a quiet map. Layer 5 says, that is not strange. If, at the moment of observation, there is a background of dozens of clusters and groups in a phase of clumping, then some light paths receive an extra dose of lens strength. In the reconstructed map, that gives a pattern of fine ripples, comparable to interference noise in a photograph.

A second effect appears in the σ8 tension, the difference between the strength of structure as inferred from the cosmic microwave background and as measured in lensing and large scale structure today. The CMB tells us how strong the ripples in the early universe were, lensing and galaxy surveys tell us how strong those ripples are now. In the standard model, a lower present day σ8 often points to a cosmos that clusters slightly less well than expected, which is interpreted as a possible indication of exotic physics or modified gravity.

In the GLV picture I add an optical component to that. The gravitational burst background acts as a kind of extra noise source in lensing measurements. Each burst adds a small, random lens component. Measured over many lines of sight and long times, that noise dampens the measurable correlations. It makes the lens field more chaotic on small scales, causing the statistical quantity we call σ8 to come out optically a little lower. The universe itself does not have to cluster less, the patterns are simply smeared out a little in our observation by the burst noise.

You can see that as a subtler form of something we know from photography. An image with very fine structures, over which you lay a grainy noise field, loses some of its sharpness. The largest contrasts remain visible, but the middle scale becomes softer, and the strength of small patterns seems to decrease. Something similar happens in lens maps when gravitational bursts dominate. The large peaks of clusters and filaments remain, but the coherence on intermediate scales is weakened. The measurement of σ8 responds to that, because it is precisely sensitive to the amplitude of structure on that middle scale.

How does Layer 5 differ from the previous layers. Layers 1 through 4 are about systematic effects, an average shortening, a directional dipole, a scale preference, a growth with depth. Layer 5 is stochastic by definition. It is not about a fixed correction, but about a background of temporary deviations. In the more technical GLV pieces, this is described as an extra noise term in the lens kernel, with its own spectrum and its own amplitude. In this book it is enough to hold on to the idea that the cosmic web is not only a static lens, but a living lens. Sometimes, for a fraction of the time, that lens is extra strong in certain places, and we integrate all of that over the full travel time of the light.

An interesting consequence is that Layer 5 plays a role in how we estimate the reliability of lens measurements. In many analyses the measurement noise is attributed to instruments, to random orientations of background galaxies, or to imperfections in data reduction. Layer 5 adds a fundamental physical noise to that. Even with an ideal instrument, even with perfect data reduction, there would be a baseline level of variation in the lens field, simply because the gravitational landscape is not static. That variation is not an error of the telescope, but a property of the universe itself.

For GLV this is not a problem, but an opportunity. It makes a testable prediction. If the gravitational burst background really exists, then in very deep lensing surveys, where instrumental noise and shape noise are well controlled, a residual must remain that you cannot remove. That residual must also have very specific behaviour, it must not be purely white noise, but must be somewhat stronger on scales where cluster mergers and halo mergers dominate. In the more technical GLV documents this is translated into a slightly flat lensing spectrum at high multipoles, and a small plateau above a certain multipole number.

Philosophically, Layer 5 once again underlines that we work with a one sided information channel. We see the universe only along the lines through which light happens to be travelling at that moment. If a cluster collapses tomorrow and causes a temporary extra lens peak, we see that peak only in the light beams passing by at that moment. The rest of the light cone is unaware of the event. The gravitational burst background is therefore also an archive of chance. A map of where the universe, during the specific period in which our light was underway, happened to be in an upswing of violence.

In relation to dark matter and dark energy, the role of Layer 5 is subtler than that of the earlier layers. Layers 1 through 4 can directly shift pieces of tension in rotation curves, mass estimates, and Hubble diagrams. Layer 5 mainly makes clear that part of what we treat as clean, deterministic signals actually sits on top of a noise floor of gravitational bursts. This makes some supposedly hard limits, for example on σ8, a little softer. The window within which the rest of GLV must fit becomes wider.

Technically, you can see Layer 5 as a small but unavoidable correction to the lens kernel. The average lens strength is described by Layers 1 through 4. The variability around that average, especially on the scales where clusters form and merge, is the domain of Layer 5. In formulas it is an extra covariance term, in words it is the recognition that the lens flickers from time to time.

Together with the first four layers, Layer 5 forms a more complete picture of the optical universe. Not only winding paths, preferred directions, scale preference, and time build up, but also a noisy, pulsing lens background that is never completely still. In the next layer one more step is added, then it is no longer only about how light is deflected, but also about how subtle differences in travel time along different paths have translated into our measurement of the cosmic clock. Where Layer 5 describes the noise in lens strength, Layer 6 describes the noise in time.