A clock map of the universe
Time runs at a different speed in every place.
Density Field Dynamics has one universal time, a single “now” shared by the whole universe. Every real clock ticks at its own rate against it, set by how deep the clock sits in the field ψ and how fast it moves through the cosmic rest frame. Here is the nearby universe drawn as a map of those clock rates, built from the measured motions of tens of thousands of galaxies.
The rule
Two things slow a clock: depth and speed.
In DFD, gravity is a field ψ that sets the refractive index of space, $n = e^{\psi}$. The field is larger where matter is packed in. A clock’s rate, measured against DFD’s universal time $t$, is
$$\frac{d\tau}{dt} \;=\; e^{-\psi/2}\,\sqrt{1-\frac{V^{2}}{c^{2}}}$$
where $V$ is the clock’s speed relative to the cosmic rest frame, the frame in which the cosmic microwave background looks the same in every direction. The first factor is depth: a clock deeper in the field runs slower. The second is speed: a clock moving through the rest frame runs slower. In the weak field of galaxies and clusters, $\psi = -2\Phi/c^{2}$, where $\Phi$ is the ordinary gravitational potential, so each clock’s fractional offset is simply
$$\text{offset} \;\approx\; \frac{\Phi}{c^{2}} \;-\; \frac{V^{2}}{2c^{2}} .$$
One part per million (ppm) is 31.6 seconds a year. The numbers are small, but they add up: they are real differences in how much time has passed.
Only comparisons
No clock can read the universal time directly. Every clock in a laboratory slows together, so nothing inside the lab notices. What shows up is the difference between clocks in different places: the gravitational redshift of light climbing out of a well, or the offsets GPS satellites correct for every day.
Round trips still read c
Nothing here changes the measured speed of light. A round trip of light reads exactly $c$ for every observer, in DFD as in relativity. DFD’s universal time and rest frame show up only in one-way comparisons, which no local experiment can make on its own.
The map
The nearby universe, by clock rate.
This slice runs through the plane that most nearby galaxies and clusters lie in (the supergalactic plane), out to about 1.1 billion light-years in each direction, with the Milky Way at the centre. Amber regions are gravity wells, where clocks run slow. Blue regions are the great empty voids, where clocks run fast. Hover or tap anywhere to read a clock; drag the depth slider to move the slice; switch on the flow to watch galaxies stream down the clock gradient.
depth 0
Colour shows how fast a clock there runs compared with a clock here: one floating at rest in the cosmic rest frame at our position, just outside the Milky Way’s pull. The clock out there moves with the local galaxy flow. The map is smoothed over about 35 million light-years, so single galaxies and cluster cores are not resolved; their own deep wells are added in the place cards below. Contours every 10 ppm. Faint particles trace the measured flow of galaxies in the slice plane. The map fades where the survey runs out of data.
In three dimensions
We are falling toward slow time.
In DFD, matter accelerates up the gradient of ψ, $\mathbf a = \tfrac{c^{2}}{2}\nabla\psi$. In clock language: things fall toward places where clocks run slow. The whole Local Group is moving at about 630 km/s through the cosmic rest frame. The map shows why. Behind us, on the side of the great empty region astronomers call the Dipole Repeller, clocks run fast. Ahead, toward the Shapley Supercluster, they run slowest. Drag to turn the cloud; scroll or pinch to zoom.
amber: slow · blue: fast · arrow: our motion
Look up, all the way out
The whole sky, out to the edge of the universe.
The map above stops about a billion light-years out, where the galaxy survey runs thin. The sky view goes the rest of the way, in three depths. Nearby shows the clocks at one distance you choose, from the galaxy-flow map. All the way through uses the Planck satellite’s map of gravitational lensing: the bending of the oldest light by everything it passed, which adds up the field ψ along each line of sight across the whole observable universe, most of it from matter billions of years back in time. The edge shows the clocks on the most distant surface we can see, the glowing gas of the early universe as it was 380,000 years after the beginning. On large angles the temperature of the cosmic microwave background is a direct reading of the depth there: cold patches are wells, where clocks ran slow, and warm patches are hills, where they ran fast.
The sky is laid out on the supergalactic plane, so the great walls of nearby galaxies run along the middle. The dashed line marks the disc of our own Milky Way, which hides the sky behind it; there the nearby map is filled in by the reconstruction, the edge map by Planck’s standard in-painting, and the lensing map is faded out.
Landmarks
How fast time runs out there.
Each card adds three things: the large-scale map, the place’s own gravity well (a galaxy or a cluster, too small for the map to resolve), and its speed through the cosmic rest frame. Every offset is against the same reference clock: one at rest in the cosmic rest frame here, just outside the Milky Way’s pull. Tap a card to find it on the map.
Close to home
Earth’s clock, term by term.
Six things set the rate of the clock on your wall. The biggest local one is not Earth or the Sun but the Milky Way: we sit inside its gravity well, and its depth at the Sun is set by the escape speed of the fastest stars, about 530 km/s. Next is our speed of 370 km/s through the cosmic rest frame, read straight off the cosmic microwave background. Earth’s own well, the one GPS corrects for, is more than a thousand times smaller than the Milky Way’s.
The time dimension
Clocks differ by place. DFD says they also differ by era.
Everything above is space: clocks in different places, all read today. DFD adds a second dimension. Its two-clock postulate says the unit of time itself was different in the early universe: before matter bound into galaxies, every microphysical rate in the hot plasma (atoms, light scattering, nuclear reactions) ran faster by one constant factor than the same rate in today’s laboratories. That factor is about 7%. It is why the expansion rate reads about 67.5 km/s/Mpc on rulers laid down by the early universe and about 72 on clocks made of today’s matter. In DFD both readings are right: one universe, read on two clocks.
The map lets us compare the two dimensions on one scale. The ladder below runs from the smallest effect to the largest; each step to the right is ten times bigger.
The gap is not in any place
To make a 7% clock difference out of depth alone, a region would need ψ offset by about 0.13. The deepest place in the map, a cluster core, reaches only a few ten-thousandths. Nowhere in the nearby universe comes within a factor of several hundred. So if the two clocks are real, the difference lives in time, between eras, not in any place.
A stated postulate
The two-clock reading is stated in the book as a postulate, and the size of the factor is measured, not derived. Deriving it is open work. The spatial map above uses only the clock law and measured galaxy motions; it does not depend on the postulate.
Looking for patterns
Tests written down before we looked.
With the whole map in hand, the next step is to look for patterns that link its layers. A map this rich will always show something if you search long enough, so the tests were written down before the final data were opened, with a strict bar: a pattern has to beat odds of 1 in 100 after allowing for how many tests were run. Nothing here is a DFD result until it passes that bar and an independent three-way check.
Read this before quoting a number
What is solid, and what is not.
Differences are solid; the overall level is not. The map pins down how much faster one place’s clocks run than another’s, so it is drawn relative to here. Two independent reconstructions of the galaxy flow, from 2023 and 2025, agree on those differences to a few ppm (Shapley against here: −63.6 and −62.1 ppm). What no survey can pin down is the level of the whole map against the universe-wide average, because structures larger than the survey lift or lower everything in it together. That level moved by about 30 ppm between the two reconstructions. It shifts every colour by the same amount and changes no difference between places, which is why nothing here is quoted against it. In DFD a constant ψ is invisible to every clock, so that level is not measurable even in principle.
The same numbers come out of general relativity. In the weak field, DFD’s clock law and Einstein’s give the same rate for a clock at a given depth and speed. Relativity treats the cosmic rest frame as a convenient choice of reference; in DFD it is physical, and so is the universal time the map is drawn against. Nothing on this page can tell the two theories apart. What DFD adds is the reading, and the time dimension in Chapter VII.
Where DFD’s extra gravity sits. Inside galaxies DFD strengthens gravity at low accelerations, the effect behind flat rotation curves. The map does not have to model it: the galaxy and cluster wells on this page come from measured motions (escape speeds, velocity dispersions, lensing masses), which already include whatever gravity is really there. On the scale of the map itself, the book’s Rest-Mass Channel postulate makes the field the ordinary Newtonian potential of the matter, which is how the map was built.
Accuracy. The map is the mean of 10,000 reconstructions of the galaxy flow from the Cosmicflows-4++ catalogue. It is most reliable within about 500 million light-years; beyond that the reconstruction’s own uncertainty grows toward the size of the flow, and the map fades. The cluster wells assume standard mass profiles and are good to about 30%. The motion term uses the flow smoothed over tens of millions of light-years; a single galaxy orbiting inside a cluster at 1,000 km/s runs another 5 ppm slow.
Who did what. Maps of the gravitational potential of the nearby universe are published work: the Cosmicflows team mapped the Laniakea supercluster, the Dipole Repeller and the flow used here. The effect of our own potential on the light we receive has been studied as a source of error in cosmology. This page applies DFD’s clock law to that published flow and draws the result as a map of clock rates.
Method and sources
- Clock law: $d\tau/dt = e^{-\psi/2}\sqrt{1-V^2/c^2}$, from the physical metric $\tilde g_{\mu\nu}=\mathrm{diag}(-c^2e^{-\psi}, e^{\psi}, e^{\psi}, e^{\psi})$ (DFD v5, sections on the formalism and on clocks). Weak field: $\Phi = -c^2\psi/2$. Linear scales: the Rest-Mass Channel postulate (appendix on the growth regime), under which the matter field clusters like cold dark matter and the potential obeys Poisson’s equation.
- Flow: Cosmicflows-4++ mean density and velocity fields over 10,000 Hamiltonian Monte Carlo steps (Courtois, Mould, Hollinger, Dupuy & Zhang 2025, A&A 701, 187; grid of 128³ cells in a 1000 Mpc/h box, version of 13 October 2026). Cross-check: the 2023 Cosmicflows-4 grouped reconstruction (Courtois et al. 2023, A&A 670, L15).
- Potential from motion: linear theory, $\mathbf v = -\tfrac{2f}{3\Omega_m H}\nabla\Phi$ with growth rate $f=\Omega_m^{0.55}$, $\Omega_m = 0.31$. Distances in Mpc/h cancel the Hubble constant out of $\Phi/c^2$. Varying $\Omega_m$ from 0.28 to 0.33 changes every depth by less than 5%.
- Our motion: the Sun moves at 369.8 km/s through the cosmic rest frame (Planck 2018 dipole); the Local Group at about 627 km/s (Kogut et al. 1993). Earth’s orbit adds a yearly swing of ±0.12 ppm.
- Milky Way well: escape speed at the Sun 528 km/s (Deason et al. 2019), plus the outer halo; Andromeda at 0.77 Mpc with about 1.5×10¹² solar masses.
- Cluster wells: central depth of a standard (NFW) profile from published masses: Virgo about 6×10¹⁴, Perseus about 7×10¹⁴, Norma, Coma and the Shapley core (A3558) about 10¹⁵ solar masses.
- The edge: Planck 2018 SMICA (no-SZ) temperature map with Planck’s in-painting of the masked plane (Planck Collaboration 2020, A&A 641, A4), smoothed to 8° and cut at multipole 30, read with the Sachs–Wolfe relation $\Phi/c^2 = 3\,\Delta T/T$ (Sachs & Wolfe 1967). Cross-check: WMAP 9-year ILC map (Bennett et al. 2013); the two agree with correlation 0.99.
- All the way through: Planck 2018 minimum-variance lensing reconstruction (Planck Collaboration 2020, A&A 641, A8), Wiener-filtered with its published noise, converted from convergence to lensing potential, multipoles 8 to 250.
- Distances are shown in light-years using the local Hubble rate of 72 km/s/Mpc.
- Context: Hoffman, Pomarède, Tully & Courtois 2017, “The Dipole Repeller”, Nature Astronomy 1, 0036; Tully et al. 2014, “The Laniakea supercluster of galaxies”, Nature 513, 71; Wojtak, Davis & Wiis 2015, “Local gravitational redshifts can bias cosmological measurements”, JCAP.
Time runs at a different rate in every place. The emptiest voids run fastest, and we are falling away from them, toward the slow clocks of Shapley.