Facebook

The Evidence for an Old Earth

The young-earth model’s claims are wanting. This essay is the positive case, strongest lines of evidence that the earth is about 4.5 billion years old and the universe about 13.8 billion, in the form a scientist would give them: what is measured, what physics connects the measurement to a time, what could go wrong, and how the result is cross-checked. It closes with the young-earth responses to each and why they have not persuaded the people who make the measurements.

The argument does not rest on any one method. It rests on the agreement of many methods that share no assumptions, run on different physics, and would fail in different ways. A single clock can be wrong. A dozen unrelated clocks that read the same time are evidence that the time is right.

1. Radioactive decay

The physics

A radioactive nucleus decays with a fixed probability per unit time, set by nuclear forces and unaffected by anything happening outside the nucleus. This has been tested directly: decay rates do not change with temperature, pressure, chemical state, gravitational field, or electromagnetic field, to a precision of parts per thousand or better, across every condition a laboratory can produce (the only known exceptions — electron-capture isotopes such as ⁷Be, and fully ionized ¹⁸⁷Re — shift by about 1% under conditions that do not occur in rocks). The half-lives of the isotopes used in dating are measured to within a fraction of a percent, and several have been measured by independent methods (direct counting, accumulation, and calibration against each other) that agree.

Two observations show the rates have been the same in the past and far away:

The Oklo natural reactor in Gabon, where uranium in a 1.7-billion-year-old ore body sustained fission for some hundreds of thousands of years. The isotope ratios left behind constrain the nuclear constants at that time to within about one part in ten million of their present values.
Supernova 1987A, 168,000 light-years away. Its light curve, after the initial flash, followed the exponential decays of ⁵⁶Co (half-life 77 days), then ⁵⁷Co (272 days), then ⁴⁴Ti (60 years), with exactly the laboratory half-lives. The decay rates in that star, at that distance and time, were the ones measured on earth. Light curves of more distant supernovae show the same.

The methods and their self-checks

The basic idea — measure parent and daughter, compute elapsed time — has an obvious weakness: one must know how much daughter was present at the start. The methods in actual use remove that assumption rather than making it:

Isochron dating measures several minerals from the same rock. Their parent/daughter ratios fall on a straight line whose slope gives the age and whose intercept gives the initial daughter content. If the rock has been disturbed, the points scatter and the method reports failure rather than a wrong age.
Uranium–lead in zircon uses two clocks in one crystal: ²³⁸U → ²⁰⁶Pb (half-life 4.47 billion years) and ²³⁵U → ²⁰⁷Pb (704 million years). Two independent decay chains with different rates must agree (“concordance”); lead loss or disturbance produces discordance that can be diagnosed and, often, corrected. Zircon excludes lead when it forms, so the initial-daughter problem is nearly absent.
Argon–argon dates are calibrated against known eruptions. The Vesuvius eruption of AD 79 was dated by ⁴⁰Ar/³⁹Ar in 1997 to 1,925 ± 94 years before the measurement — that is, to within a century of the historical date, on a 2,000-year-old sample, with a 1.25-billion-year half-life.
Independent systems with different chemistry — U–Pb, Rb–Sr, Sm–Nd, K–Ar, Lu–Hf, Re–Os — are applied to the same rocks. Their parents and daughters behave differently under heat, water, and melting, so a process that reset one would not reset the others by the same amount. They agree.

The results

Meteorites, the leftover material of the solar system’s formation, give 4.567 billion years (from calcium-aluminum inclusions dated by Pb–Pb, confirmed by Al–Mg, Hf–W, and other short-lived systems). This is the age of the solar system, known to about a million years.
The oldest earth minerals — zircons from the Jack Hills of Western Australia — are 4.4 billion years. The oldest rocks (Acasta gneiss, Canada) are 4.03 billion. The Moon’s crust is 4.4–4.5 billion.
The geological column has been dated at hundreds of thousands of points: volcanic ash beds between sedimentary layers, lava flows, intrusions. The dates increase downward without exception across the whole column, from the Quaternary to the Precambrian, and the Sauk-to-Tejas interval of the seminar’s Flood lectures spans about 540 million years.

The young-earth response

The RATE project (ICR and Creation Research Society, 1997–2005) accepted that the measured amounts of decay products are real — that the rocks contain the daughter isotopes that billions of years of decay at present rates would produce — and proposed that decay rates were accelerated by a factor of about a billion during creation week and the Flood. Two consequences follow, both acknowledged in the RATE authors’ own volume as unsolved. The heat released by 4.5 billion years’ worth of uranium, thorium, and potassium decay compressed into a year is on the order of 10³¹ joules, enough to melt the crust hundreds of times over and comparable to the energy needed to melt the entire mantle. And accelerated decay would have delivered lethal radiation doses to everything alive, including the occupants of the Ark. RATE also reported helium retained in zircons and radiocarbon in diamonds and coal as evidence of youth; the helium result depends on diffusion-rate and temperature assumptions that have been contested in detail, and the radiocarbon measurements (about 0.1% of modern levels) are at the level of instrument background and of in-situ production by uranium decay in the surrounding rock. Neither result explains why every other system gives concordant old ages.

2. Counting years directly

Radioactivity is one clock. Nature also keeps calendars — layers laid down one per year — and these can simply be counted, then checked against each other and against radiocarbon.

Tree rings. Cross-dated, overlapping sequences of living and dead trees give continuous, year-by-year chronologies: the bristlecone pine record about 9,000 years, the German oak-pine record about 12,500 years, combined with others to nearly 14,000. Each ring is dated to the year. Radiocarbon measured in the rings tracks the ring count.
Lake varves. Lake Suigetsu in Japan has about 52,000 annual couplets, cross-checked against radiocarbon over the whole span; Lake Van, the Cariaco Basin, and others give comparable records. The Eocene Green River Formation of Wyoming, within the seminar’s “receding Flood” sequence, contains several million couplets with seasonal chemistry.
Ice cores. Greenland cores have been counted layer by layer to about 110,000 years, each year marked by seasonal cycles in dust, isotopes, and acidity, with the sulfate of historically known eruptions (Laki 1783, Tambora 1815, Vesuvius AD 79, Thera in the second millennium BC) at the expected depths. Antarctic cores reach 800,000 years.
Speleothems. Cave stalagmites deposit annual laminae and can be dated independently by uranium–thorium; the two agree, and the records extend to 500,000 years.
Coral. Corals lay down daily and annual growth lines. Devonian corals, about 380 million years old on the radiometric scale, show about 400 daily lines per annual band. The earth’s rotation is slowing by about 2 milliseconds per century from tidal friction (measured by atomic clocks, and confirmed by ancient eclipse records); projecting that rate back 380 million years gives about 400 days in the year. Tidal rhythmites — sediment laminae recording daily and monthly tides — from 620-million-year-old rocks give about 13 lunar months and 400 days per year, again matching. The lunar laser-ranging measurement of the Moon receding at 3.8 cm/year is the same physics seen from the other end.

Each of these calendars is independent of radioactive decay. Where they overlap with radiocarbon or uranium–thorium, they agree; where they overlap with each other, they agree.

3. Rates measured today, distances measured today

A third family of arguments needs no isotopes and no counting: measure how fast something is happening, measure how far it has gone, and divide.

Seafloor spreading. GPS measures the plates moving at 1–10 cm per year today. The Atlantic is about 4,500 km wide and spreads at about 2.5 cm per year, giving 180 million years — which is the radiometric age of the oldest Atlantic crust, drilled off the American and African coasts. Magnetic stripes on the seafloor, symmetric about the ridges, record the earth’s field reversals; their spacing matches the spreading rate and the reversal chronology dated on land. Sediment thickness on the seafloor increases with distance from the ridges at the rate at which sediment is observed to accumulate. The depth of the ocean floor increases with the square root of its age, exactly as the cooling of a conducting plate predicts over 100 million years.
Hotspot tracks. The Hawaiian-Emperor chain runs 6,000 km from the active volcanoes of Hawaii to the 80-million-year-old Meiji seamount. Radiometric ages along the chain increase linearly with distance at about 9 cm per year — the speed at which GPS shows the Pacific plate moving now. The Yellowstone track across the Snake River Plain gives the same result for North America over 16 million years.
Erosion and deposition. Rates of erosion measured by cosmogenic nuclides (tens of meters per million years for most landscapes), of pelagic sediment accumulation (millimeters per thousand years), of salt deposition, reef growth, peat accumulation, and soil formation all give the durations of the deposits they produce, and those durations — thousands to millions of years per formation — match the radiometric ages of the formations.
The Earth’s heat. Continents have a measured temperature gradient and a measured heat flow; the crust’s radioactivity supplies about half of it. A cooling and heat-producing body of the earth’s composition reaches the observed thermal state only after billions of years.
Impact craters. The Moon’s surfaces carry crater densities that increase with the radiometric ages of the Apollo samples collected from them; the relation is used to date other planetary surfaces. Saturation cratering on the oldest lunar highlands requires billions of years at any impact rate consistent with what is observed today.

4. Astronomy

Distance and light travel

The distance to a star or galaxy is measured by methods that overlap and calibrate each other: parallax (geometric, now to tens of thousands of light-years with Gaia), Cepheid variables, Type Ia supernovae, and, for a few objects, pure geometry. Supernova 1987A is the best geometric case: the light from the explosion illuminated a ring of gas around the star, and the delay between the flash and the ring’s brightening, combined with the ring’s angular size, gives the distance — about 168,000 light-years — with no ladder at all. Light from Andromeda has traveled 2.5 million years; from the farthest galaxies, more than 13 billion.

The young-earth literature offers two answers. One (Humphreys) proposes that gravitational time dilation during creation week let billions of years pass in the cosmos while days passed on earth; the model has been criticized within creationism (Conner and Page, 1998) for not producing the required effect, and it predicts distortions in the light of distant objects that are not seen. The other (Lisle) adopts a convention in which light travels toward earth instantaneously. That convention is not contradicted by any experiment — the one-way speed of light cannot be measured without a synchronization convention — but it is not supported by one either, and it requires that every distant event we see, including the supernova decay curves that show earth-like physics 168,000 light-years away, be occurring now, with physics everywhere arranged to look as though it had a past.

The ages of stars

The Sun. Helioseismology — measuring the Sun’s internal structure from its surface oscillations — gives an age of 4.57 billion years from the amount of helium accumulated in its core, independently matching the meteorites.
Star clusters. Stars burn hydrogen for a time set by their mass; a cluster’s oldest surviving main-sequence stars give its age. Globular clusters give 12–13 billion years.
White dwarfs. A white dwarf is a dead star cooling at a rate set by known physics. The coolest, faintest white dwarfs in the Galactic disk have been cooling for about 9–10 billion years; those in the halo, longer.
Radioactive stars. Uranium and thorium lines in the spectra of very old halo stars give decay ages of 12–14 billion years — the same nuclear physics as terrestrial dating, applied at astronomical distance.

The universe

The expansion of the universe, measured through redshift and distance, gives the time since the expansion began: 13.8 billion years, with an uncertainty of about 1%. Three independent observations confirm the picture of a hot, dense early universe: the cosmic microwave background, a blackbody at 2.725 K with the temperature fluctuations a hot early universe predicts; the abundances of hydrogen, deuterium, helium, and lithium, which match the nuclear reactions of the first three minutes; and the direct measurement of the background’s temperature in the past — absorption lines in gas clouds at high redshift show the radiation was hotter then, by exactly the factor (1 + z) that expansion predicts. The universe has a measured thermal history, and it is billions of years long.

5. Genetics

Living things carry their own clock. Mutations accumulate in DNA at a rate that is now measured directly by sequencing parents and children: about 1.2 × 10⁻⁸ per site per generation in humans, roughly 60–70 new mutations per person.

Human diversity. Any two unrelated people differ at about three million sites. If all humans descended from the eight people on the Ark about 200 generations ago, each lineage could have accumulated about 3 × 10⁹ × 1.2 × 10⁻⁸ × 200 ≈ 7,000 new differences — two lineages, about 14,000 — against three million observed. The young-earth reply is that Adam and Eve were created with built-in diversity; but four founding chromosomes can carry at most four versions of any gene, and the eight survivors of the Flood at most ten, whereas the HLA-B gene alone has thousands of versions differing from each other by dozens of accumulated mutations. The observed diversity requires roughly a million years of accumulation in a population of about ten thousand, which is what the fossil and archaeological record independently give.
The molecular clock. Differences between human and chimpanzee DNA, at the measured mutation rate, give a divergence of about 6–7 million years — the same as the radiometric ages of the earliest hominin fossils, dated by argon–argon on volcanic ash. The two clocks share no physics.
Ancient DNA has been recovered and sequenced from bones and sediments dated to more than a million years, showing the expected accumulated differences from their living relatives.

6. Why the agreement is the argument

Set the clocks side by side. Uranium–lead, rubidium–strontium, samarium–neodymium, potassium–argon, and rhenium–osmium share no chemistry. Tree rings, varves, ice layers, and coral bands share nothing with any of them. Seafloor spreading, hotspot tracks, and lunar recession are measured with GPS, magnetometers, and lasers. Stellar ages come from nuclear fusion and cooling physics; the universe’s from expansion and thermodynamics; genetic ages from sequenced pedigrees.

Each of these could, in principle, be wrong. For the earth to be 6,000 years old, every one of them would have to be wrong, by a factor of between a thousand and a million, in the same direction, by the same amount where they overlap — and the error would have to leave no trace, since the young-earth model’s own researchers have had fifty years and the same laboratories to find it. The only mechanism proposed, accelerated decay, releases enough heat to melt the planet and does nothing to explain the tree rings, the ice cores, the seafloor, the corals, the stars, or the genomes, none of which involves decay.

This is the puddle of wax from the seminar’s candle. A catastrophe that reset one clock would leave the others reading the old time, and the discrepancy would be the evidence. The clocks agree. That agreement is not an assumption that nothing happened; it is the measurement that shows nothing did.

7. The theological footnote

None of this addresses whether God created the world; it addresses when. The people who built this evidence include Steno, Buckland, and Sedgwick in geology and Lemaître in cosmology, believing men who took it as a description of God’s work. The young-earth model asks that the evidence be a false appearance — light created in transit, isotopes created in decayed proportion, a Bronze Age that left harbors at the wrong sea level. The alternative, held by most Christians since the 1830s, is that the heavens declare the glory of God in the language of what is actually there.

260907 – Evidence for an Old Earth