The Monolith on Phobos
In 1998, the Mars Global Surveyor took several photographs of Mars satellite Phobos with its high-resolution camera. On seeing this image some time ago, I noticed what appeared to be an unusual formation in the processed version of image SP255103 that I downloaded from the MGS Archive maintained by the United States Geological Survey (USGS). My interpretation, however, was that while it was interesting, it could probably be explained as a rounded boulder casting a long shadow due to a low sun elevation.
Mr. Efrain Palermo had also examined this image and came to a conclusion considerably different than my own. He had noticed something of possible significance that I had missed on my cursory examination: the white streak extending down from the left side of the object (or out from its base along the ground, in my interpretation). To him, this suggested that what I saw as a "dome-like rock" embedded in the surface of the moon was in fact only the "roof" of a vertical tower. The adjoining white streak, in his interpretation, was a small part of the vertical side of this tower illuminated by sunlight, the rest of it being hidden in the object's own shadow.
We discussed the image via email. I objected that the problem with his interpretation was that the orientation of the long axes of the crater ellipses in the image suggested strongly that the MGS took the image from almost a horizontal position relative to the moon's surface. It seemed to me that the camera was to the left relative to the image and looking towards the right across the moon's surface, with the horizon somewhere past the image's right edge. Any tall structure should be oriented in the direction of the short axes of the crater ellipses in an image taken from a near-horizontal perspective. In this case, it seemed that a tower should extend from the left of the image to the right, not from the bottom toward the top as Efrain proposed. The image as Efrain and I first saw it is shown in Figure 1.
Figure 1. Section from processed JPEG image.
To make sure that the spacecraft viewing angle was as low as it appears in the image, I referred to the image data on the USGS archive of MGS images. The data for SP255103 is at:
I was surprised to find that the viewing angle was not low at all. The "emission angle," which is in the thirteenth column of each image listing, was 19.84 degrees. The emission angle is the angle off the vertical direction of the planet surface to the camera, so the spacecraft was looking almost straight down from above, at an elevation angle of 90 minus 19.84 degrees or 70.16 degrees.
But why are the craters so elongated? It is because the image we both were looking at was a processed JPEG version we had downloaded from the browser page for the image at: http://ida.wr.usgs.gov/html/orb_0551/55103.html.
The craters are deceptively elongated in the processed image because it was stretched for some reason. The raw image shows nearly circular craters, as would be expected for the low emission angle.
Efrain's interpretation of the object is shown on the version of the image below that I had corrected by stretching the image until the craters were circular rather than elliptical. His interpretation of the object is shown in Figure 2.
Figure 2. Orthorectified version of the section from SP255103 with the "monolith."
The raw image is also available on the browser page, and it turns out my "orthorectification" was unnecessary. The section from the raw image is shown in Figure 3. The contrast has been adjusted but the image dimensions have not been altered from the raw image in any way. It is virtually identical to my orthorectified version. This should be very close to how the object actually appeared to the camera.
Figure 3. Section from raw image after contrast adjustment.
There seems to be no good reason to put the word "monolith" in quotes to describe this object. Efrain's interpretation is absolutely correct. At the very least, this object is a monolithic block of stone, although it's high reflectivity may indicate that it is composed of something else. But how likely is it that it's artificial? That depends in part on how tall the object is relative to the width of its base. A block of stone several times longer than its height created from the impact of some large meteoroid would be unlikely to land on its narrow end and remain upright. The greater the height-to-width ratio, the less likely it is that this monolith was placed on the surface of Phobos by the forces of Nature. On this question, the MGS image data has another surprise: the incidence angle (in the fourteenth column of each image listing) for the principal (center) point in the photograph is 43.47 degrees. The incidence angle is the angle between the sun and the vertical direction relative to the surface being imaged. The height of an object can be simply computed as the length of its shadow divided by the tangent of the incidence angle. The length of the shadow is 53 pixels and the width of the object is 15. The tangent of the 43.47 degree incidence angle is 0.948, implying a height-to-width ratio of 53 / (15 * .948). This would make the height of the object 3.7 times its width. These are the proportions of a skyscraper, not of a natural object. And with the stated image resolution of 2.4 meters per pixels, the object would be the height of a forty-two-story office building!
But before getting carried away with the notion of an immense building rearing up from the surface of Mars' moon Phobos, a little cautious consideration of what the image data really means is in order. Determinations of an object's height from shadow length and sun angle depend upon knowing the sun's elevation above the surface on which the shadow is cast. Unfortunately, the sun elevation for spacecraft images only tells us the sun's elevation at a point on a perfect sphere centered on the targeted planet or moon. And Phobos is far from a perfect sphere. Its dimensions are 27 X 21.6X 18.8 kilometers. It is 40% wider at its greatest width than it is at its most narrow width.
Even for a world that is spherical in shape, there can be considerable variations in local topography - hills and valleys - that make the implications of shadow length uncertain. This was a problem with the "Blair Cuspid" objects on the Earth's moon. However, there are ways to estimate the true sun angle on a sloping surface from image brightness variations. These methods are based on the knowledge that the moon is, on average, a nearly perfect sphere. These methods, as applied to the Blair Cuspids are described at:
(NOTE: Further review of MGS image data indicates that the non-spherical shape of Phobos was in fact accounted for in the computation of the stated sun elevation. See 2/15/00 update).
But there is another clue besides the shadow length that suggests this monolith on Phobos has an appreciable height: the craters are nearly circular, indicating the spacecraft was looking almost straight down at the surface at this point in the photograph. Yet the vertical edge of the object (outlined in red in Figure 2) is still 14 pixels long. If the object is upright relative to the surrounding surface, it would still be twice the height of the apparent width of its base assuming that the emission (viewing) angle of the camera was the stated 19.84 degrees at this point in the photograph. Although assumptions about the actual camera viewing angle are uncertain for essentially the same reasons that assumptions about the sun angle and shadow length are, it does provide an estimate of height independent of the estimate based on the shadow length. Both ways of estimating indicate an extremely unusual object may stand on the surface of Phobos.
Figure 4 shows Efrain's idea of what this object might look like viewed from the surface. While I think it may be more vertical and with a sloping "roof," rather than leaning to one side as he does, his mock-up seems a reasonable and perhaps even a conservative depiction. The dimensions he gives the object make its height twice the width of its base (only about 28 stories high rather than 42 as implied by the shadow length). It could well be much higher.
Figure 4. Mock-up of the possible appearance of the monolith as viewed from the surface of Phobos.
Despite the uncertainties in these estimates of the object's shape, it clearly is unusual enough to warrant another look. A single image taken from the nearly horizontal perspective in Figure 4 would be most informative. Unfortunately, the Mars Global Surveyor, now in a low mapping orbit around Mars, will never have another close encounter with Phobos. That will have to remain the task of some future spacecraft. Given NASA's history of avoiding any intentional examination of anomalous objects, it will probably have to happen by accident. But Phobos is a very small world, so if other spacecraft have close encounters with Phobos in the future, there is a fair chance we eventually will see this monolith again.
--- Lan Fleming
Efrain would be interested in any comments or questions on his discovery. He can be contacted at:
See second image and some of the other unusual objects from the first image in the 2/15/00 update.