Planck: separate an early-Universe signal from the foreground sky

Planck launched in May 2009 alongside Herschel, but its central objective was different: measuring the cosmic microwave background, or CMB, across the sky. This radiation carries information from the time the Universe became transparent to freely traveling light, roughly 380,000 years after the Big Bang. It is not a photograph of the instant the Universe began.

Tiny variations in the radiation’s temperature and polarization contain information about early structure and later propagation. Scientists compare those patterns with predictions of cosmological models. The spacecraft does not directly display the Universe’s age, composition or expansion history on a detector; those quantities are inferred by analyzing the observations within a stated model.

Planck’s Low Frequency Instrument and High Frequency Instrument together covered nine frequency bands. Different foreground components, including emission from our Galaxy, have different spectral behavior. Multiple bands help distinguish those contributions from the CMB instead of treating every feature in a raw microwave map as an early-Universe signal.

Instrument calibration, detector noise, angular response and foreground modeling all influence the resulting products. Removing a foreground is an analysis step with uncertainty, not a cosmetic edit made to produce a smoother picture. The cleanroom photograph shows Planck during preparation and does not contain the microwave measurements or their later component separation.

A sky map records or estimates a quantity as a function of direction. Statistical summaries describe how patterns vary with angular scale, and model comparisons connect those summaries with physical parameters. A fitted value should be reported with its uncertainty and assumptions. Agreement with a model is evidence to evaluate, rather than a guarantee that every possible alternative has been eliminated.

For a fictional comparison between two cosmological estimates, check whether they use the same observations, foreground treatment and model. A numerical difference can arise from any of those choices as well as from measurement uncertainty. Planck’s legacy includes the maps, calibration and analysis documentation that let researchers revisit these relationships and test how robust a conclusion is to the assumptions behind it.