01A spectrum is a pattern across wavelengths#
Spectroscopy studies how light changes across its wavelengths. A spectrum may be shown as a coloured image or as a graph of brightness against wavelength. Matter can emit, absorb, transmit, or reflect light differently at particular wavelengths, so a spectrum can carry clues about what an object contains and its conditions. It adds information that a picture alone usually cannot show.[1][2]
Astronomers use these patterns to investigate stars, gas clouds, planets, and galaxies. For example, spectra of exoplanet atmospheres can show wavelength-dependent absorption in starlight that has passed through an atmosphere. Such a measurement is evidence to interpret, rather than a direct sample of the world.[1][2]
A spectrum represents measured intensity or flux as a function of wavelength, frequency, or photon energy; plots must be read with their axis direction, scale, and units in mind. A smooth continuum can be modified by absorption or emission features. Line positions and shapes, continuum shape, and relative feature strengths encode different information, but depend on temperature, composition, density, motion, and radiative transfer through the source.[7]
In transit spectroscopy, researchers compare a star’s measured spectrum during a planetary transit with measurements outside transit. Wavelength-dependent changes can constrain atmospheric opacity and an effective transit radius. Clouds, hazes, temperature structure, stellar surface variation, and instrument effects can produce overlapping signals, so retrieved atmospheric properties depend on assumptions and joint modelling.[1][2]
02Instruments separate and record the light#
A telescope gathers light and sends it to a spectrograph, which separates wavelengths using an element such as a prism or diffraction grating. A detector records the dispersed light. The James Webb Space Telescope carries several spectroscopic modes, while other observatories use instruments designed for different wavelength ranges and kinds of targets.[2][3]
The instrument does not simply produce a ready-made answer. Observers and processing software correct detector effects, identify wavelengths, remove or estimate background light, extract a spectrum, and calibrate its brightness. Those steps help turn detector measurements into quantities that can be compared and scientifically interpreted.[5][6]
Spectrographs disperse incident light onto detector pixels, and the wavelength solution maps detector position to wavelength. Resolving power is commonly expressed as R = λ/Δλ, where λ is wavelength and Δλ is the width of a resolvable feature near it; R is dimensionless and depends on the instrument mode and measurement convention. Higher resolving power can separate closer features, but spreads photons over more detector elements and does not guarantee adequate signal-to-noise.[2][3]
Calibration uses instrument models and reference observations to map pixels to wavelength, correct detector and optical response, estimate background, and convert counts to flux when appropriate. In the JWST pipeline, wavelength coordinates use instrument-specific reference files and pointing information, while flux conversion and background handling vary by mode. The exact processing sequence is instrument-specific; calibrated products still carry uncertainties and data-quality flags.[5]
03Interpretation depends on calibration and context#
A feature in a spectrum can suggest a chemical or physical process, but researchers compare it with laboratory measurements and models before drawing conclusions. Instrument sensitivity, background light, limited resolution, noise, and the conditions of the source all affect the result. Spectroscopy is powerful because it makes distant measurements possible, but it does not remove uncertainty.[1][8]
Uncertainty includes photon and detector noise as well as systematic effects. STScI identifies sensitivity calibration, aperture throughput, source centering, extraction size, and background subtraction among factors that can limit flux accuracy; wavelength accuracy also depends on the dispersion solution, wavelength zero point, and source position in the slit. Their magnitude varies with instrument, configuration, target, and reduction choices, so a calibration figure for one mode must not be generalized to all spectroscopy.[8]
Scientific interpretation therefore requires propagated measurement errors, inspection of data-quality information, and checks for systematic effects. Blended lines may be unresolved, different physical conditions may produce similar line ratios, and uncertain source models can leave multiple explanations consistent with a spectrum. For an exoplanet atmosphere, a molecular signature constrains atmospheric models; it does not establish biological activity without additional, contextual evidence.[6][8][4]
