How we know the colors of Mars

By Don Davis


The beginning of my interest in planetary color was an article on the subject by Andrew T. Young in the May 1985 issue of Sky and Telescope, pages 399-403. I then corresponded with Andrew and he sent me graphs and color chips referenced to the Munsell Book of Color. He included a set of Munsell measurements he made of sample copies of the article with corrections on what the colors should be! He even surveyed various copies to try to quantify the variations in the print run. Based on his helpful information and consideration of the brightness values, or albedos, of the planets I put together a web page displaying attempts to show planets in relation to measured values.

The Martian surface has long been seen to be divided into lighter 'deserts', darker roughly equatorial regions, and seasonally variable polar caps. Through the big refracting telescopes used in many early observations Mars appeared as a light orange globe with blue green darker regions. This impression was partly based on effects of the chromatic aberration of such telescopes, which tended to make the focusing dependent on the colors of the subject. A less focused image from the other end of the spectrum will 'spill over' the darker portions of the overall brighter parts of the planet which is at the redder end of the visible spectrum. Later visual studies by Gerard Kuiper using carefully daylight adapted vision and color chips made the dark areas appear more neutral in color, brown to 'moss green' in the darkest spots. Telescopic observers have long used red filters to emphasize the surface contrasts, and blue filters to bring out clouds across the planet.

Modern reflecting telescopes coupled with advances in cameras have allowed dedicated amateurs to obtain images that eluded even the greatest observatories in past decades. Film exposures needed time to capture a relatively dim subject, during which the turbulent motion of the image as seen through the atmosphere would build up a blurred result. Digital imaging methods allow very short exposures to be made, each less degraded to begin with, and when combined with many other similar exposures the 'areas of agreement' add up. This dramatically realizes the potential of telescopes. One location with the good fortune of commonly steady atmospheric conditions is the Pic Du Midi Observatory in the French Alps. Even in the days dominated by film their photos were second to none in detail, but with modern imaging and processing techniques their Mars images in particular show how far ground based observations have advanced.

Close approaches of Mars to Earth have also allowed Hubble to obtain its best views of the planet:

The outbound Rosetta spacecraft obtained color images of Mars in 2007, they were made with the excellent OSIRIS camera using RGB color filters:

Mars has had color cameras of varying types placed in orbit, making maps as well as obtaining many close ups. The first 'color' camera at Mars orbit was that of the Viking Orbiter, which had a color filter wheel. A rapid photo sequence using a grayscale camera with a color filter wheel allowing records of the red, green and blue light in a scene is a time proven method of obtaining high fidelity color images. Unfortunately the filters used in Viking were marginal for color work, although dedicated image processors have made fine color products. Recent examples of these have appeared in, where many planetary photo gurus share their processing of related images, sharing useful info on factors affecting how such pictures appear.
Later orbiters had better cameras, generally abandoning 'snapshot' type cameras for 'pushbroom' devices which capture a 'line' of the subject below, rapidly building up a scanned image as the orbiter passes over the planet. Mars Odyssey used a different camera system capturing red, green and violet light as well as infra red. Odyssey could capture fine color differences across the scene below but color balancing them to something resembling RGB images was more problematical, mainly due to light leaks in the violet channel. I took it upon myself to manually remove such light leaks from some earlier images, the results were accepted by the imaging team and included in their daily releases here:

Color images from orbiters often show dark dune fields as vivid blue. Much of this is probably due to a thin cloud cover contributing disproportionately to that color channel especially in the darker portions. This has caused areas that are dark to give misleading colors in the initial image product. It is also understandably instructive to emphasize the color differences between dune, dust, and rock covered regions, leading to a tendency to deliberately exaggerate the color differences rather than to record what someones visual impression might be. Such reservations about the color in an image often appears in official captions. The Mars Reconnaissance Orbiter has been gathering color and grayscale images of unprecedented detail and quality. The related image site is a rewarding treasure trove of images.

India's Mars Orbiter Mission has a decent color camera that has obtained numerous global views. The Planetary Society has a page on images from this mission, with the data to download:


Mars as seen from the surface by landers and rovers


For the surface of Mars, we are presented with several independent observations of the colors in the sky and scenery. Although further exploration continues to show increasing variety in rock types, the planet wide fine dust and the sky appear virtually the same wherever on Mars we have looked. References below to summaries of the data show the cameras of various missions are seeing similarly colored things.

The actual surface colors were first measured by the Viking landers and presented in JGR Vol 82, no. 28 September 30, 1977 pages 4401-4411 'Spectrophotometric and color estimates of the Viking Lander sites'.

The next lander, Pathfinder, delivered consistent results with its different camera system, as documented in JGR volume 104, issue E4 April 25 1999 pages 8781-8794

The MER rovers have had much color work done with their cameras, a fine example is 'Chromaticity of the Martian sky as observed by the Mars Exploration Rover PANCAM instruments Jim Bell and others JGR planets Volume 111 issue E12 december 2006.

A delightful page on the MER Rover PANCAM camera color images can be seen here:

The PANCAM page dedicated to detailed aspects of determining the true colors of Mars is here:

Here also is an explanation for the variability of color saturation and hue seen on the PANCAM photo pages:

'The images on these pages are created without an accurate luminance scaling factor calculation, and will therefore sometimes fail to accurately represent the true luminance of their subjects. Because of this, certain images will appear to be too bright or too "orange," while others may appear to be too dark or too "brown." Future work on our algorithms may allow us to add luminance calculations like those initially described for Pancam work by Bell et al. (2006), hopefully improving the accuracy of such images.'

Infra-red filters are usually used by the MER rovers as a substitute for red, in order to tease out more of the color contrasts and mineralogical differences in the scenery. The sundial color chips look very different in infra red and visible, which attracted attention on line when initially released. When seen in RGB color, the sundial color patches can be seen initially bright and then being coated with time by the steadily settling dust.
The Mars Science Rover 'Curiosity' has color sensor cameras which have been used to gather views of the scenery and the hardware. These cameras capture the surroundings in color, and the hues in the scenery are showing the different cameras are seeing the same thing. Most Mars surface photography is exposed to provide maximum contrast in the surface details, in effect moderately overexposing the scene in absolute terms for the sake of obtaining the most pictorial information on the inherently dark surroundings.
A sample of Mars dust resembles cumin powder in color, somewhat lighter than the average planetary albedo of 0.16. One day when samples are brought back their darkness in relation to surface photography will seem surprising to many.
The overall measured brightness values on Mars from Earth, from orbit and from the surface tell us the same thing. The relative RGB values are such that the Green channel will be a grayscale version of the scene. With the red channel the scene is brighter (and with higher contrast between dust covered areas and the basaltic rocks) and the blue channel is darker and with less contrast between surface types. Besides the spectral response graphs for the camera hardware we have color calibration targets on the landers and rovers. The gray scales on them are particularly valuable for reconstructing color values. In studio photography color charts are often placed in the frame under the same lighting to allow a standard object to be compared with a similar example when preparing a copy.
Viking had color and grayscale charts, the latter more useful as things turned out The blue pigments used in that mission also had a tendency to photograph as purple. Pathfinder had a color target, photographed in the foreground before the camera mast was raised. Phoenix had basic color charts as well, with its camera showing a similarly colored scene as elsewhere.\

The Mars Rovers use a sundial / color chart with tiny mirrors also giving a glimpse of the sky, at least before the dust covered them. Early in the missions all these charts complimented the camera telemetry and allowed particularly good comparisons to be made between them and the Martian surface colors.

This is the basis of my web page presenting early MER rover color images:

In this and other examples I am letting the color chart 'drive' the color and contrast tuning. Using the sunlit color chart derived background colors, I assume wider views of the same scene in similar lighting will continue the observed color trends. Look at the ground and rocks in the background of the chart photo, which reasonably resembles the ground sundial photos. The Martian surface seen in that and the wider view are very similar. The Opportunity rover landed in a different kind of surface but the dust color there and elsewhere on Mars seems to be uniform.
Such matching of the scene to the color chart of course assumes a 'white' sunlight lighting as on Earth. The sky contributes a significant amount to the ambient lighting, generally of the tan ochre of the brighter parts near the horizon. This should be taken into account to better match the color of the lighting which will greatly influence an initial visual impression. The sky glow is brighter and bluer near the Sun due to a narrow 'forward scattering' effect of the fine dust. Similar effects can be seen in terrestrial dust storms. In Egypt I saw and photographed a version of such effects caused by significant blowing dust, the sky tan colored near the horizon and light blue nearest the Sun. Elsewhere in the Martian sky the sunlit dust appears as a luminous version of the color of the dust covering much of the surface, brighter and more saturated near the horizon. Near the zenith the skies appear a dark gray.
The basaltic rocks seen ever since the Viking missions are a dark gray color, with the dust settling and forming ephemeral coatings of variable thickness and shape from the whims of the winds. Upon occasion the specular highlights from such dark rocks can reflect the blueish atmospheric glare near the Sun, shifting and bringing striking apparent color changes to some rock surfaces over the hours with changing Sun angle. Such effects on wide regions of flat rocky surfaces have been seen from orbit as a moving bright zone. This effect may have even been seen from Earth.
There have been rare instances where my appraisal of how things would look varies from what is officially published by the related space agencies. But I also remind myself that their purpose is rarely to try to duplicate what human eyes would see, it is to show the information the best way possible no matter how much brighter or dimmer or beyond the visible spectrum the real thing may be. Image presentation related issues like exposure and color values are a matter of variables and judgment.

However it is clear the revealed appearance of Martian scenery and skies is consistent enough between the surface locations seen so far as to build up an impression that can closely approach a visual impression of a possible eventual human visitor.
What will that impression be? As soon as one leaves the artificially lit spacecraft interior the scene would drastically change in average color. An amber tan hue would initially dominate, with the bright dust laden sky near the horizon adding to the scene a lighter luminous version of the color of the fine dust likely to be everywhere. Rocks still show their colors, in many places the dark gray of basalts. In a short time one would 'get used' to the color shift and probably do a fine job of picking out subtle color differences.

The dust extends tens of miles above the ground, dimming the Sun significantly as it nears the horizon. Directly overhead the minimum dust is looked through, there the sky is a darker gray. The Moons and a few brighter planets may be seen in the skies. Magically removing all the suspended dust would reveal a sky something like that seen from a balloon 100,000 feet up, essentially black except for a horizon hugging blue glow.
At night Mars would be a terrible place for astronomy, with all but the bright stars being lost in the murk as one looks away from the zenith. Efforts to photograph the Earth rise have run up against the dust obscuring it. Earth rises as a Morning Star ahead of the Sun, which is steadily brightening that part of the sky at the same time Earth is rising into visibility through less dust. A luminous blue glow surrounds the tiny Sun, more visible from behind the obscuring horizon dust as it climbs in the sky.