Stellar Luminosity and Radius: Stefan–Boltzmann Law with Worked Examples

How do radius and temperature determine luminosity?

For a spherical star, the effective-temperature relation is L=4πR²σT_eff⁴. Luminosity L is total radiative power, radius R supplies the emitting area, and σ is the Stefan–Boltzmann constant. At fixed radius, doubling temperature multiplies luminosity by sixteen. At fixed temperature, doubling radius multiplies it by four.

The received flux at a distant observer is a separate quantity. For isotropic emission without intervening absorption, f_bol=L/(4πd²). Distance changes how bright the source appears; it does not change the source's intrinsic luminosity. All numerical stars in this guide are constructed examples, not measurements of named objects.

This relationship extends the previous discussion of a spectrum's peak. A peak wavelength can suggest a temperature under a blackbody model, but temperature alone cannot supply the total power of a star with unknown size.

Separate surface flux, luminosity and observed flux

The two fluxes share units but refer to different surfaces. F_surface is evaluated at the emitting surface. f_bol is evaluated where the radiation is received. Their equality is not implied by having the same unit. For a distant observer, radiation has spread over a much larger sphere.

Bolometric means integrated over the whole spectrum. A reading through a visible-light filter is a band flux and cannot automatically replace f_bol. Hot and cool sources place different fractions of their emission in that band, so a common visual brightness need not imply a common bolometric luminosity.

Spectral radiance has an additional per-solid-angle factor and often a per-wavelength or per-frequency factor. Integrating blackbody radiance over frequency gives σT⁴/π; the outward hemispheric flux is σT⁴. This distinction explains the factor of π that sometimes appears when comparing a software model with the surface-flux formula.

Use kelvin and state the solar conversion constants

The calculations use σ=5.670374419×10⁻⁸ W m⁻² K⁻⁴, the displayed digits of the NIST constant. Temperatures must be absolute temperatures in kelvin. Applying the fourth power to degrees Celsius is a unit error, even if the resulting number looks plausible.

The nominal solar conversion factors are Rᴺ☉=6.957×10⁸ m and Lᴺ☉=3.828×10²⁶ W. IAU Resolution B3 defines these factors exactly for consistent conversion. They should not be described as measurements of an unchanging physical Sun. Here the superscript N marks their nominal status.

The separately adopted nominal solar effective temperature is 5772 K. Combining the displayed σ with the nominal radius and luminosity gives approximately 5772.003429 K. That tiny difference is a consequence of rounded, separately adopted constants. For the tables, calculate in SI and divide by Lᴺ☉; do not force all nominal quantities to satisfy an extra exact identity.

A label such as two solar radii also refers to radius, not diameter. If the input is a diameter, halve it first. Squaring a diameter in place of a radius overestimates the area and luminosity by four.

Work through a 6000 K example in SI units

Take a hypothetical star with R=6.957×10⁸ m and T_eff=6000 K. First compute the outward power per square metre: σT⁴≈7.348805247×10⁷ W m⁻². Its spherical area is 4πR²≈6.082104402×10¹⁸ m².

Multiplying those quantities gives L≈4.469620074×10²⁶ W. Dividing by the nominal solar luminosity gives approximately 1.167612350 Lᴺ☉. The area supplies square metres that cancel the inverse square metres in the surface flux, leaving watts. That dimensional check should precede a comparison with any catalogue.

The sphere's total emitting area is 4πR². The projected circular area πR² answers a different geometrical question and would underestimate this luminosity by four. A planet intercepting light from a star uses its projected area, which is one reason the two areas are easily confused.

Compare radius and temperature independently

The first three rows have the same surface flux because their temperature is the same. Their total luminosities differ because their areas differ. Compare the 3000 K and 6000 K rows at the same radius: the temperature doubles and the luminosity increases by sixteen.

A hotter object is therefore not necessarily more luminous than every cooler one. A sufficiently large cooler object can radiate more total power than a small hotter object. Temperature describes the surface energy output per area; radius determines how much area contributes.

The displayed digits are useful for reproducing the calculation. They do not imply that real stellar radii and temperatures are known to nine decimal places. Measurement precision and arithmetic precision are separate, and a final observational result should use the uncertainty of its inputs.

Read the three fixed-radius curves

The vertical axis is logarithmic: equal vertical steps represent equal ratios, not equal added watts. At every temperature, the two-radius curve is four times the one-radius curve, while the half-radius curve is one quarter of it. The plotted values have not been normalized to a common peak.

These are comparisons at constant radius, not the path that an individual star follows through its life. Radius can change during stellar evolution. This plot also puts increasing temperature to the right; a conventional Hertzsprung–Russell diagram often reverses that temperature direction. Read the actual axes before interpreting a line.

Use ratios to see the degeneracy

For two spherical sources described by the same relation, L₂/L₁=(R₂/R₁)²(T₂/T₁)⁴. The physical constant cancels. Ratios are particularly helpful for testing an intuition before doing a long calculation in watts.

The final row keeps luminosity unchanged: a star four times the radius but half the temperature has 16 times the area and one sixteenth the surface flux. Equal total power therefore does not identify a unique radius or temperature. An additional independent constraint is required.

Taking logarithms makes the trade-off explicit: log L equals 2 log R plus 4 log T and a constant. At fixed luminosity, radius scales as temperature to the power −2. A temperature estimate that rises by ten percent changes the inferred radius by a factor 1/1.1², approximately 0.826446, when luminosity is held fixed.

Infer radius when luminosity and temperature are known

Rearrange the relation to R=√[L/(4πσT_eff⁴)]. Convert a luminosity quoted in nominal solar units to watts before applying the SI formula. The result is in metres; divide by Rᴺ☉ only afterward if a solar-unit answer is desired.

At one nominal solar luminosity and 3000 K, the radius is about 3.701780398 nominal solar radii. Increasing the temperature to 6000 K at the same luminosity makes the radius four times smaller. This inverse calculation provides a separate check of the ratio reasoning.

Solving for temperature instead gives T_eff=[L/(4πσR²)]¹ᐟ⁴. Neither rearrangement creates new information. If both luminosity and radius are unknown, one received-flux measurement does not determine them together; distance and another independent constraint are still needed.

Convert intrinsic power into received flux

For the following exercise set L exactly to one nominal solar luminosity and use 1 au=149597870700 m. Assume isotropic emission, a distance much larger than the source radius, and no intervening extinction. The entries represent flux at a surface perpendicular to the incoming rays, not an average over a planet.

The computed 1 au value is approximately 1361.166465 W m⁻². The IAU's separately rounded nominal solar irradiance is 1361 W m⁻². The difference again reflects conversion conventions; the table is an inverse-square calculation from Lᴺ☉, not a new measurement of sunlight.

At twice the distance the flux is one quarter; at ten times the distance it is one hundredth. Reversing the calculation gives L=4πd²f_bol. Extinction, incomplete spectral coverage, an unresolved companion or an unsuitable isotropy assumption can bias the inferred luminosity. Those issues cannot be repaired by adding more decimal places.

Propagate uncertainty with the stated assumptions

For small independent uncertainties in radius and effective temperature, the approximate relative luminosity uncertainty is √[(2σ_R/R)²+(4σ_T/T)²]. For radius inferred from independent luminosity and temperature, it is √[(σ_L/(2L))²+(2σ_T/T)²]. Here σ with a subscript means an input uncertainty, while the un-subscripted σ in the radiation law is the physical constant.

Temperature's fourth-power influence makes a modest temperature error consequential. The percentages in the table are standard uncertainty propagation under the independence assumption. They are not worst-case bounds and do not by themselves specify a confidence interval.

If luminosity and temperature came from the same fitted data, their errors may be correlated. Include the covariance rather than blindly applying the independent-input expression. Similarly, recalculating luminosity from an assumed radius and then using it to “measure” that same radius is a circular consistency check, not an independent observation.

What effective temperature does and does not mean

Effective temperature is defined by matching the total outward surface flux to σT_eff⁴. A real star can have absorption lines and a non-Planckian spectrum while still having a useful effective temperature. The equation for total luminosity therefore does not assert that every wavelength follows a perfect blackbody curve.

Colour temperature, brightness temperature and an interior or core temperature answer other questions. A temperature fitted from one spectral region should not be silently substituted for an effective temperature without checking the model. A nonuniform or rapidly rotating source can also require a more careful treatment of surface area and temperature distribution.

Read Wien's law for peak-based reasoning, magnitudes and flux for brightness comparisons, and parallax and distance uncertainty for a separate distance constraint. The practical sequence is to identify the measured quantity, state the model, carry units through the calculation, and report the uncertainty with the result.

Frequently asked questions