The ‘Face’ on Mars

Cydonia Mensae

The part of Mars that has attracted most attention since satellite exploration of the planet began is the Cydonia Mensae region, located in the planet’s northern hemisphere. The region bears traces of former shorelines – it is now accepted by most astronomers that Mars was once home to large bodies of water – and of eroded hills and mountains, corresponding to terrestrial buttes and mesas. A number of features were seen on Viking Orbiter photographs from the mid 1970s that led some researchers to suggest that they were artificial in origin. One of the most impressive parts of this claim was the concentration of anomalies in a single area; these anomalies included what was claimed to be an enormous carved face, three- and five-sided pyramids showing bilateral symmetry and walled compounds. Moreover, some of these features appeared to be in alignment with each other. The so-called ‘Face’ has attracted most attention.

Viking Orbiter 1 Frame 070a13
taken 24 August 1976

The ‘Face’

The ‘Face’ has been known as that since NASA engineers dubbed it such as a convenient shorthand. Located at about 40.9º N 9.45º W, it was photographed by the Viking Orbiter 1 on 25 July 1976 whilst obtaining images of a site suitable for Viking Lander 2. Eighteen images of the region were taken, but of these, eleven have resolutions worse than 550 m per pixel (in other words, no feature less than 550 m across can be seen on these images). Only two of them (frames 070A13, taken on 24 August 1976, and 035A72, taken on 20 July 1976) have resolutions of better than 50 m per pixel. They show a subrectangular elevated feature that displays roughly bilateral symmetry similar to a human face. The length is estimated as 2 km, the width as 1.5 km and the height as about 400 m. The long axis of the face is aligned at a bearing of approximately 329º to the Martian North Pole. In Frame 070A13, the sun is 28º above the horizon, while in 035A72, it is 11º; these differences in angle reveal slightly different details in the ‘Face’.

Detail from Viking Orbiter 1 Frame 035a72
taken on 20 July 1976

The ‘Face’ lies close to the boundary between an area of flat-topped mesas and conical hills to its southwest, and a much flatter region in which it lies. Claims have been made for the presence of a former shoreline close to the ‘face’, which may be important in considering its likely origin. Erosion is generally cited as the main factor in the creation of the landforms in this region, but a variety of claims were made during the 1980s and 90s that this was not sufficient to explain a number of features of the ‘Face’. The original view of astronomers was that the northern past of Mars had originally been covered with a sediment that subsequently eroded, perhaps as a result of wind action. As evidence for the former presence of surface water on the planet has increased, so it has begun to seem more likely that the erosion in this region may have been exacerbated by former seas.

The ‘Face’

The initial announcement of the discovery of this feature dismissed it as being “formed by shadows giving the illusion of eyes, nose and mouth”. The public took little notice of it until DiPietro and Molenaar published an analysis of the formation in 1982, in which they concluded that it was artificial. Since then, a number of popular writers have taken up the theme, most notably Richard Hoagland and Graham Hancock. The widespread availability of computers for processing digital images from the mid-1980s onwards led to a great deal of amateur work on the Viking Orbiter photographs. Different techniques for enhancing the photographs have been used, including techniques that attempt to deduce three-dimensional shape from shadows, the interpolation of pixels and so on. These techniques were all used to make the ‘Face’ appear more detailed and more humanoid than on the original satellite images. The ‘Face’ has become an iconic image, familiar to millions; its ghost-like features stare blankly from the covers of numerous books and videos. Demands for photographs of higher quality than those of 1976 – with their resolutions of no better than 43 m per pixel – grew during the 1990s, especially when NASA announced the forthcoming launch of Mars Observer on 22 September 1992. Equipped with a camera that could achieve a best resolution of up to 1.4 m per pixel, the probe was lost on 21 August 1993, three days before entering Martian orbit.

A new image of the face obtained in 1998
by Mars Orbital Camera,
Frame SPO-1-220/03

Predictably, the loss of the spacecraft prompted suggestions of conspiracy. They can be discounted on the grounds of cost alone: had NASA wished to reassure the public that it was interested in finding out more about the face whilst being determined to prove it natural, it could have found a less expensive way of doing it than sabotaging a mission designed to obtain better photographs. Indeed, the less expensive means is what the conspiracy theorists suggested about the results of the next probe, the Mars Global Surveyor: they declared the first photograph of the ‘Face’ obtained by the Mars Orbital Camera (Frame SPO-1-220/03) on 5 April 1998 to be fake. The original release of the photograph in an unenhanced form ought to have reassured the public that NASA was being as fair as possible.

An enhanced version of
the 1998 image

This prompted yet more claims of unfair play when Malin Space Systems (which had designed the camera) released an enhanced version on the following day, removing the more egregious sampling errors from the raw data. The enhanced version shows none of the signs of artificiality alleged for the 1970s images, despite the claims of its supporters. The 1998 image, unlike the Viking Orbiter images, is an oblique, with all the distortion that entails. Even so, the ‘Face’ appears as a highly eroded ridge, with a steep perimeter. A second photograph from the Mars Orbital Camera, taken on 3 June 2000 and released on 31 January 2001, shows only a small part of the ‘Face’, but it confirms the 1998 photograph in every respect. This one, by contrast, is a near-vertical image. Supporters of artificiality have subsequently focused on the bilateral symmetry of the object, with its central ridge (the nose) and the apparent valley below the nose (the mouth); there is, however, no sign of a depression to the northwest of the southwestern end of the ridge that would provide the missing ‘left eye’, nor does the ridge appear to be central to the mesa, confirming the lack of symmetry on the unprocessed version of Viking Orbiter image 070A13. To explain the lack of resemblance to the original view of the ‘face’ as a highly structured and artificial carving, they now have to accept either that it has suffered a great deal of erosion from its pristine form or that NASA has tampered with the photograph taken by the Mars Orbital Camera, as it was the identification of the pristine form that led to the claims for artificiality being made in the first place.

The ‘Face’ therefore emerges as an optical illusion (exactly as its discoverers in 1976 had claimed), known technically as pareidolia. This is the phenomenon that allows believers to see an image of Mother Theresa in a cinnamon bun or the Arabic name of Allah in a sliced aubergine. The basis of the illusion is the human brain’s tendency to make understandable and detailed patterns from vague stimuli; the same ability allows us to recognise melodies from short or distorted fragments and to see numbers in patterns of dots used for tests of colour blindness. Faces are one of the first patterns the infant human learns to recognise, so it is unsurprising that a face-like mesa on Mars should be 'read'; by so many as an actual representation of a human (or closely humanoid) face.

This page was last updated on 28 July 2007
Written by Keith Fitzpatrick-Matthews