A background needs more than one measured brightness
In 1990, the COBE satellite team presented a spectrum of the cosmic microwave background that closely followed a single thermal curve. The instrument, FIRAS, had compared the sky with a carefully made cold blackbody. It was a particularly direct question to ask of the hot-universe picture: does the old radiation have the same spectral shape as matter and light in thermal equilibrium?
A background can be detected at one radio frequency without establishing its entire spectrum. Several unrelated sources could contribute at that frequency. Measuring many frequencies adds a much stronger requirement. One temperature must account for the changing brightness throughout the observed band. The COBE result connected the sky’s faint microwave glow to a physical state once familiar to any hot, opaque environment: radiation repeatedly exchanging energy with matter.
The satellite carried a piece of the comparison experiment
COBE launched in November 1989 with three instruments. FIRAS, the Far Infrared Absolute Spectrophotometer, measured the spectrum, while the separate Differential Microwave Radiometers mapped temperature differences between directions. A spacecraft placed the experiment above Earth’s emitting and absorbing atmosphere. The challenge then became comparing an extraordinarily cold sky with an equally well understood reference.
FIRAS used a differential design. It compared radiation entering from the sky with radiation from an internal reference, looking for the small difference between them. An external blackbody calibrator could also be placed across the sky aperture. Its absorbing structure was designed to return very little unwanted reflection, and thermometers measured its temperature. This let the team test the instrument against a source whose thermal emission they could calculate. The calibrator’s physical temperature anchored the radiation measurement.
A blackbody fixes the whole curve once its temperature is chosen
A blackbody is an ideal absorber and emitter in thermal equilibrium. Its brightness at each frequency follows Planck’s law. Raise the temperature and the radiation grows stronger, with its broad maximum moving to higher frequency. Choose a temperature and the law fixes the entire curve; the brightness in each channel is not free to move independently.
That makes the test demanding. A mixture of arbitrary radio sources would have to mimic this particular shape over a wide band. In a hot early universe, repeated interactions naturally bring radiation toward a thermal distribution. As the universe expands, all wavelengths stretch together. A blackbody spectrum keeps its shape under this stretching, with a progressively lower temperature. The cold thermal curve therefore carries information about a much hotter past.
The sky also contains emission from material closer to home. Warm Galactic dust contributes particularly at the higher-frequency end. The team compared sky regions and modelled such foreground contributions before asking how closely the remaining cosmic component followed a blackbody. An error in a foreground can leave a frequency-dependent residual, which is why a striking curve still needs careful calibration and separation of emissions.
Small residuals made the thermal test precise
The first FIRAS report gave a temperature of 2.735 kelvin with an uncertainty of 0.06 kelvin. The 1996 full-data analysis by Dale Fixsen and colleagues reported root-mean-square departures below fifty parts per million of the background spectrum’s peak brightness. That statement specifies the size of the residual across the spectrum relative to the peak. It does not mean every individual channel or every calibration parameter was known to fifty parts per million.
Later work refined the temperature scale. Fixsen’s 2009 review combined temperature measurements to obtain 2.72548 kelvin with an uncertainty of 0.00057 kelvin. The review’s value and the early FIRAS value belong to different stages of the measurement history. The lasting result is the exceptionally thermal shape. Limits on distortions can then constrain how much energy was added to the radiation after the early processes that could restore thermal equilibrium.
A spectrum and a map answered complementary questions
The 2006 Nobel Prize recognized John Mather and George Smoot for the blackbody form and anisotropy of the microwave background. Mather’s contribution centred on the spectral measurement; Smoot’s on the tiny direction-dependent differences measured by DMR. The two findings fit together. An almost thermal glow explains the common background temperature. Slight variations across the sky give cosmologists a way to study the departures from perfect uniformity.
FIRAS set a standard for connecting a cosmic claim to a physical comparison. The experiment carried a calibrated thermal source, placed it in the same measurement path as the sky and examined the remaining difference. The universe’s history became accessible through a carefully controlled question about what a cold object should emit.