Good day, student of introductory astronomy!

Since you were interested enough to inquire, I have decided to pass along the following to you. This is in regards to our discussions of the concepts of light, color, and thermal (also known as blackbody) radiation.

The first concept to keep straight in your mind is the difference between reflected light and emitted light. We observe our surroundings through the reflection of visible light, emitted by our Sun or man-made sources. Emitted light is composed of photons generated by the matter emitting the light - it's an intrinsic source of light. Reflected light consists of photons whose origin is elsewhere but reflected (scattered) from some object. Light falling upon matter can be absorbed, reflected, or transmitted (meaning that the light and matter interact very little - the matter is transparent), or often some combination.

Absorption, emission, reflection, transmission

A photon of light that is absorbed by an atom or molecule must have had a particular energy that was capable of internally ``exciting'' that atom or molecule. Once absorbed, the photon has given up ALL of its energy to the atom or molecule, and so ceases to exist. For example, the electrons bound to the nuclei of atoms can be excited to higher energy orbitals through the absorption of photons. The electrons of molecules can also be excited to higher energy orbitals (molecules can also be excited into modes of vibration and rotation). If that excited atom or molecule can (spontaneously) de-excite itself before colliding with another particle of matter, a photon will be emitted. The emitted photon will have an energy corresponding to the how much energy the atom or molecule lost in the process of de-excitation. The corresponding wavelength of light emitted is inversely proportional to the energy of that photon.

Each atom (e.g., hydrogen, carbon, oxygen) and molecule (e.g., H2O, CO2) of a given elemental composition has a unique set of these internal energy levels (or modes of excitation). Thus we can identify the elements and molecules in some material simply by studying the wavelengths of light that are absorbed or emitted.

If the light passing near the atom or molecule hasn't the right energy to be absorbed, there is a chance that it might just scatter (or bounce) off one of the bound electrons, sending the photon off in some semi-random direction. If the photon doesn't interact much with the matter, then it will just pass on through - and we say that the light is transmitted through the matter*.


Color is our brain's interpretation of our eyes' detection of the different wavelengths of visible light. This is the regime of light, with wavelengths between 400 nanometers and 700 nanometers (1 nanometer = 1 billionth of a meter = 10-9 meter), that our eyes are sensitive to.  Our eyes have 3 ``color'' sensitive detectors called cones: red, green, and blue. Each type of cone is sensitive to a broad range in wavelength (called a response curve) of visible light, but this sensitivity is roughly bell-shaped, peaking at a particular wavelength, with lower sensitivity for wavelengths to either side of the peak. These response curves may be seen here. Our perception of color comes from the brain's interpretation of how many photons of particular wavelengths fall within the response curves of each of the 3 types of cones. For example, an equal mixture1 of red and green light is interpreted as yellow, equal mixture of red and blue light is interpreted as magenta, and equal mixture of blue and green light is interpreted as cyan. Red, green, and blue are called primary colors; magenta, yellow, and cyan are called secondary colors. Our brain interprets a roughly equal mixture of red, green, and blue light (or a roughly equal mixture of secondary color light) as white light. Other ``colors'' are possible with the mixture of primary and/or secondary color light with white light. For example, pink is the combination of yellow and magenta light, or equivalently due to the combination of red and white light. It is probably more accurate to speak of degrees of response between the red, green & blue cones, rather than mixtures of red, green & blue light as I described above. Go here and here for more information. Further discussion of color rendering of spectra can be found here.

Because the cones' sensitivities overlap in wavelength, any colored source of light will appear white if it is intense enough (all three cone types will be fully activated). It is also true that the cones are insensitive to low light intensities; the least sensitive blue cones stop reacting first, followed by the red ones, then the green ones, as the intensity of light diminishes. At low light levels, another detector in our retina takes over - the rods - and these are ``color blind,'' and their response curve is also shown here. So very dim sources of broad spectrum light will appear dull white or grey.

The color of reflected light

Note that so far we are discussing the mixing of different color light, which is an additive process, not the mixing of dyes, pigments, or filters which is a subtractive process. The molecules in dyes and pigments absorb light strongly at some wavelengths and reflect others; colored filters absorb strongly at some wavelengths and transmit others. When white light falls upon an object that absorbs blue (i.e., shorter wavelength visible) light strongly, the color yellow results. An object that absorbs strongly red (i.e., longer wavelength visible) light, reflects a cyan colored spectrum. This is the reason for the colors of Uranus and Neptune, which have methane in their atmospheres that strongly absorbs longer wavelength visible (and infrared) light. Chlorophyll is a molecule that absorbs both short and long wavelength visible light - a green dominated spectrum is what is reflected, and accounts for the color of leaves. Something that appears ``black'' absorbs visible light of all forms very effectively.

There is nothing special or fundamental about color. It is merely our body's physiological (eye-brain) reaction to the detection of light of various wavelengths between 400 nm and 700 nm.

Color and thermal radiation, stars

As the temperature of a hunk of metal in a blacksmith's fire is increased, the heated metal follows Wien's and Stefan's laws of thermal radiators. Wien's law says that the wavelength of the peak intensity in the broad spectrum of light emitted will shift to shorter wavelengths; Stefan's law says that the total amount of light emitted per square meter of its surface per second increases. Initially, the metal is at room temperature and radiates with a thermal radiation spectrum that peaks at about 10 microns (1 micron is 10-6m = 1000 nm). It is invisible to the human eye in a dark room. As the temperature increases, the thermal radiation spectrum's maximum (or peak) intensity shifts to shorter wavelength infrared light, and our skin and nervous system first detect the radiation emitted from the warming metal when we place our hand in its vicinity. When the metal's temperature approaches 700 Kelvin or so, our eyes begin seeing a dim, red glow (in an otherwise dark room). As the temperature further increases, the metal glows ever more brightly and its ``color''2 changes from red, to orange, to yellow, to yellow-white, to nearly white. The change in color is due to the thermal radiation spectrum shifting to shorter wavelengths with the increase in temperature. Ultimately, the object becomes ``white hot'' because light from the broad, thermal radiation spectrum is emitted roughly uniformly across the visible spectrum and has equally activated the red, green, and blue cones - sometimes to the extent of saturating them (which the brain also interprets as white).

When you look at the brighter stars at night, look closely and you will discern that stars appear to have slightly different, albeit subtle, colors! To my eye, the brighter stars can appear pale blue-white, white, pale yellow-white, pale yellow-orange to pale orange.3 This is a real effect! Your eye is crudely measuring the temperature of the star - the above color sequence is one of decreasing temperature. Stars are, to a good approximation, thermal (blackbody) radiators. However, keep in mind that the apparent subtle colors of the planets is not a temperature effect. The visible light from planets is due to sunlight reflected from their atmospheres or surfaces, and any coloration you perceive is due to the absorption of certain wavelengths of light, as discussed above. Planets do emit thermal radiation-like spectra due to their finite temperatures (due to the absorption of sunlight, as well as internal energy sources), but these appear in the infrared portion of the electromagnetic spectrum of light.

Color perception is actually a rather complex phenomenon, and there is much more to it than meets the eye (or than we have discussed here). However, I hope to have provided you with a flavor of what goes on.

thanks for your interest,

Professor Korista

*A technical detail (and so for only for the really curious). A low density gas' interaction with light differs in detail from that of bulk matter's interaction with light of a given wavelength. Reflection and refraction could (or should) actually be thought of as phenomena of coherent scattering.  Thus, for example, light traversing all the way through transparent glass has still interacted with the electrons in the glass. But this light didn't have the correct energy to cause a change in the quantum state of the electrons, and so the interaction time is insignificant.

1More acurately: an equal response from the red and green cones due to the absorption rate of photons of the appropriate energies (or wavelengths).

 2The colors of "thermal radiators" are never stark or "saturated" (such as a red traffic light), because visible blackbody radiation spans the full visible spectrum, even if it favors long over short wavelength, for example.

3Even the brighter stars in the night sky are dim enough that the color-blind rods are contributing significantly to your perception of them.

Kirk Korista
Professor of Astronomy
Department of Physics
Western Michigan University
last edited: 15 February 2010