Which early epochs require uncertain extrapolation

A successful test has a domain

The near-blackbody cosmic radiation spectrum and light-element abundance inference provide major tests of thermal-history models. They involve different observations, calibration issues, reaction physics and model assumptions.

These records do not directly display every earlier temperature, particle species or high-energy mechanism. Extending a successful description to hotter stages requires stating what continues unchanged and what new physics might matter.

A classical expansion equation can be evaluated outside the interval where the assumed contents have been tested. That mathematical continuation is a conditional prediction, rather than proof that those contents existed at every extrapolated stage.

The following supplied extra-fluid example isolates that logical issue. It neither proposes an observationally accepted early history nor gives a current constraint on an actual particle or field.

Keep the common radiation normalization explicit

Consider two ideal spatially flat general-relativistic backgrounds sharing a separately conserved radiation bath with constant thermal degrees of freedom and no entropy exchange. Put x = T/T_c = a_c/a over 1 ≤ x ≤ 100; no physical value of T_c is assigned.

Background A contains radiation alone. Background B contains the same radiation plus a positive separately conserved stiff fluid with pressure P_s = ε_s, where ε_s is energy density. No microscopic origin for that fluid is asserted.

The stiff-fluid conservation law gives ε_s proportional to a⁻⁶, while radiation scales as a⁻⁴. Choose ε_s(T_c)/ε_r(T_c) = 10⁻⁴. At a higher bath temperature ratio x, their energy-density ratio is therefore 10⁻⁴x².

Let H_star be the radiation-only rate of A at T_c. Then H_A = H_star x² and H_B/H_A = √(1 + 10⁻⁴x²). H_star is not the total B rate at T_c; B is already slightly faster there.

A small reference contribution can grow backward

At x = 1, the extra-to-radiation ratio is 0.0001 and H_B/H_A ≈ 1.000050. The fractional rate difference is about 0.000050, or 0.005 percent. It is small under this chosen normalization.

At x = 10, the energy ratio is 0.01 and the rate ratio is about 1.004988, a 0.499 percent increase. At x = 100, the components have equal energy density and the rate ratio is √2 ≈ 1.414214.

Equal extra and radiation energy doubles the total B energy relative to A at that temperature, but it increases the rate by √2 rather than 2. The square root follows from the flat Friedmann relation between H² and total energy density.

Increasing x means looking toward a hotter, smaller-scale-factor stage in this constructed comparison. It does not specify an observed date, a particle threshold or an allowed thermal regime for the actual Universe.

Numerical closeness does not establish an empirical fit

Choose a 0.1 percent tolerance solely as a teaching comparison at x = 1. The two supplied rates differ by less than that chosen value there. The tolerance is not an observed error bar, confidence interval or limit from a real experiment.

The calculation shows that reference-stage closeness under that chosen comparison does not force the same contents or rates at much higher temperatures. It does not show that the B amplitude passes all actual abundance, cosmological or laboratory constraints.

An actual alternative must be tested against all applicable evidence, including how its extra expansion affects reaction histories and later observables. One selected rate comparison cannot replace that assessment.

The component’s assumed equation of state and separate conservation are also part of the extrapolation. A physical model might change its behavior or transfer entropy; this constructed example excludes those possibilities over its declared finite interval.

Separate observations, inference and uncertain extension

A useful thermal-history statement identifies the measured record, the physical model relating it to an epoch, and the assumptions extending the account earlier. Different steps can have different uncertainty and evidence strength.

Light-element tests, for example, combine later abundance measurements with reaction inputs and an expansion history. Agreement within a specified analysis does not certify every higher-temperature species or provide a direct observation of a physical origin.

The same discipline applies to hypothetical high-energy transitions, reheating mechanisms or extra contents. Their mathematical consequences can be explored, while their empirical status must be established separately.

The next chapter develops nucleosynthesis and its measurements more closely. Here the key skill is to state a conditional continuation accurately: preserve what is tested, identify what is extrapolated and ask which observable consequences could distinguish alternatives.

Worked reasoning

Try the reasoning

1. Does the chosen 0.1 percent tolerance establish a measured confidence interval?

   Show the answer and reasoning

   No.

   It is a synthetic comparison threshold with no observational dataset attached.

2. Is H_star the total B rate at T_c?

   Show the answer and reasoning

   No.

   It is the radiation-only A rate; B has H_B(T_c) = H_star √(1 + 10⁻⁴).

3. Does the example establish that an extra stiff component existed or that its chosen amplitude is observationally allowed?

   Show the answer and reasoning

   No.

   It illustrates a conditional extrapolation. An actual model must satisfy the full applicable constraints.

What to carry forward

• Thermal-history evidence tests specific physical accounts over specified domains.
• A component small at one reference stage can grow important in a conditional hotter-stage extrapolation.
• A chosen tolerance and an internally consistent toy calculation do not establish an empirical fit or an observed origin.