Select Language

Cart Cart

Advanced Search
Solar System
Media and Links




updated from Meta Research Bulletin 1#4:41-46 (1992)

Tom Van Flandern <>


Abstract. Interlocking evidence implies that the largest feature on the surface of the planet Mars is likely to have been formed by impacts from the tidal decay of a former band of equatorial moons of Mars similar to Phobos and Deimos.


The largest contiguous feature on the surface of the planet Mars is a huge canyon named Valles Marineris (VM for short). Extending about 5000 km in length near the equator of Mars, this intriguing system spans nearly one-fourth of the way around the planet. (See Figure 1 and Figure 2. Spacecraft images for this article are courtesy of NASA/JPL/MSSS.) The main part of the canyon ranges from 2-8 km deep, and 50-100 km wide, with walls sloping upward from a fairly flat bottom over about a 10-km span on either side. At its east end the canyon spreads out into an amorphous, broad, flat "mouth" that eventually blends into the surrounding terrain. Several troughs, both inside and nearby to the main canyon, run generally parallel to it. One notable trough is one about half the length of the main canyon and lies parallel to it just to the north. Inside are occasional stretches of parallel subcanyons forming an extended network. No single subcanyon is longer than 800 km. The steep vertical sides of the canyons often have "sculpted" edges or "wrinkle ridges." [See additional pictures in: Abell, G.O., Morrison, D., and Wolff, S.C., Exploration of the Universe, Saunders College Publishing, Philadelphia (1991); data from: R.A. Schultz, "Structural development of Corprates Chasma and Western Ophir Planum, Valles Marineris, Mars", JGR-Planets 96:22,777-22,792 (1991).]


Figure 1. Global map of Mars showing only the highest and deepest features.


The standard explanation is that the canyons of VM were formed by tectonic processes – Mars-quakes. It seems reasonably certain that they were not water-carved like the Grand Canyon on Earth, since they are not sinuous (meandering), and have no tributaries or outlets. From their somewhat "boxy" appearance, and for want of a better explanation, they are believed to be fissures produced by uplift resulting from crustal tensions, as opposed to collapse features. The presence of three of the highest volcanoes on the surface of Mars to the west, where uplift must have been operating, further reinforces this notion. The presence of some visible landslides and infilling in the canyons seems consistent with the quake hypothesis.


The theory that Valles Marineris originated from tectonic activity on Mars is a typical product of inductive reasoning -- using the observations to make an educated guess of their cause. The author has argued that such a process is haphazard at best [Dark Matter, Missing Planets and New Comets, North Atlantic Books (1993)]. If a starting point is available that allows deductive reasoning, and if those deductions explain the observations a priori without contradiction, and if they make testable predictions the failure of which will falsify the hypothesis, such a theory has a far stronger likelihood of faithfully representing nature.


Figure 2. Valles Marineris, with two of the Tharsis volcanoes at the left.


In the case of the origin of VM, a different starting point is suggested by other lines of evidence. The previous reference cited also discusses the origin of the two asteroidal moons of Mars, Phobos and Deimos, from gravitational screen capture. For present purposes, we note simply the evidence presented there for a former larger population of asteroidal moonlets in short-term orbits around Mars. But a single roughly Phobos-sized moonlet would suffice to apply deductive methodology to what happens when such an object has its orbit decay into the Martian atmosphere.


Following well-known dynamical processes, mutual collisions among numerous moonlets tend to damp out both inclination to the equator and eccentricity. As a consequence, the longest-surviving members of a larger population of moonlets will tend to have direct, low-eccentricity orbits near the equator of Mars, just as Phobos and Deimos do today. Atmospheric drag and tidal forces eventually cause most of these objects to decay from orbit and burn up in the thin Martian atmosphere or hit the surface. Observations suggest that Phobos is in the process now of doing the same thing. Its predicted impact on Mars is about 40 million years in the future.


Consistent with this starting point, evidence has been reported of an unusually large number of objects that have struck the Martian surface in the past at angles of less than 15 degrees [P.H. Schultz and D.A. Crawford, "Martian impacts and Phobos' grooves," Sci.News 136:334 (1989)]. Numerous low-angle impacts very nicely fit into this scenario calling for the decay of many former asteroidal moonlets, and are difficult to explain in other ways.


Let us consider the details of the fate of one large moonlet as it undergoes its demise toward Mars. A moonlet of appreciable mass will accelerate its decay because of the drag of atmospheric friction, but will be unable to burn up in the very thin Martian atmosphere, which has only about 1% of the density of Earth's atmosphere. Instead of causing the mass to burn up, the atmospheric drag will tend to slowly lower the moonlet's already near-circular orbit, just as happens for artificial satellites of the Earth. The object may continue to orbit within the atmosphere for a number of revolutions, depending on its mass to cross-sectional area ratio. On each orbit its mean height above the surface drops progressively.


            To be a bit more quantitative, to dispose of all the orbital angular momentum that a decaying moonlet has, it must encounter a total mass of atmosphere comparable to its own mass. In a single orbit of Mars, the atmospheric mass encountered is equal to the cross-sectional area of the moonlet multiplied by the circumference of the orbit and the density of the atmosphere at that height. The moonlet mass is its cross-sectional area multiplied by 4r/3 and the moonlet density, where r is the moonlet mean radius. To a good approximation, the moonlet will not decay until the ratio of the orbit circumference to the moonlet radius multiplied by the ratio of atmospheric density to moonlet density becomes of order unity. Near the Martian surface, the latter ratio is less that 10-5 for typical moonlet densities of about 2 g/cc. The Martian circumference is about 2x104 km. So a moonlet must have a radius of at least a few hundred meters to survive atmospheric entry for a full revolution at low altitudes. To avoid slowing so much that it can no longer stay in orbit requires not losing more than about 10% of its speed, of a moonlet of a few kilometers radius.


So we have a significant mass spiraling very slowly downward in a prograde orbit near the Martian equator. What must happen next is quite predictable. The moonlet will eventually make contact with someplace on the Martian surface. Specifically, it must be the highest surface location in its orbit plane near the Martian equator. That highest equatorial place would be in the Tharsis region to the immediate west of VM, where three large, high shield volcanoes are found in a roughly north-south line across the equator. There is almost no other choice – any slowly decaying, equatorial moonlet must first touch the Martian surface in the vicinity of one of those volcanoes, whose height exceeds anything else in the equatorial band.


The moonlet still has an enormous amount of angular momentum from its orbital speed. So this initial grazing contact does not produce an explosion or a crater. But it would rob the moonlet of a bit of its angular momentum, such that the moonlet would not be able to maintain orbital speed or altitude for the rest of that particular orbit of the planet. At some point down range a thousand kilometers or so, the moonlet's lowered orbit will cause it to burrow into the surface tangentially. What follows next may be described as a body with the mass of many mountains barreling along parallel to the surface at a couple of kilometers per second, faster than a bullet, and digging in as it goes. It would drag, then roll, and perhaps eventually tunnel across the Martian surface until its considerable momentum was brought to a halt, probably fragmenting along the way. In short, it would gouge out a canyon and end up shattered and buried.


Moreover, if there were more than one such moonlet, each would suffer a similar fate, and might even plow into the same canyon, widening and deepening it. Other moonlets might produce parallel, nearby canyons. Such a series of events may have been the origin of Valles Marineris on Mars. Note that, while the existence of past Martian moonlets to form VM is inferred indirectly, it takes no inferences to predict confidently that the large inner moon Phobos will eventually suffer just this fate. One day in about 40 million years, Phobos will decay in the Martian atmosphere, graze across the Tharsis volcanoes, and plow into Valles Marineris, widening and deepening the canyon yet again. And if one cause for producing Valles-Marineris-like canyons is known to be already operating, Occam’s razor argues that we should not invent another.


We already note some other strong features of this hypothesis. The decayed-moonlet origin for VM predicts that VM will lie near the Martian equator and roughly parallel to it, that it was formed in a prograde direction, and that it must lie somewhat to the east of the highest equatorial region on the planet's surface. The Mars-quake hypothesis simply accepts a canyon anywhere as resulting from a chance fault in the planet's crust, and needs to suggest that the specific location (a few thousand kilometers east of the Tharsis volcanoes) and the direction of VM (west to east along the equator) are random.


Figure 3. Sculpting of Valles Marineris canyon walls.


Here are some additional quantitative details. An equatorial satellite must have an orbital velocity of 3.56 km/s at 3400 km from the center of Mars (the planet's radius), and an orbital period of 1.67 hours. Mars' surface spins in the same direction at 0.24 km/s, so the relative velocity between moonlet and surface is 3.32 km/s. (A typical bullet travels about 1 km/s.) An orbiting satellite would rotate with the same face always toward Mars because of tidal forces. So its own spin speed would be negligible prior to its first contact with the surface. Hence the impactor must skid until it can rotate fast enough to roll. Visual evidence in the canyons is consistent with this picture, with the extreme west (starting) end of the canyon looking like the smooth excavation expected to accompany first contact of the moonlet as it de-orbits. As the roll continues, friction dissipates energy until the impactor "digs in" and fissures or explodes, with its remaining energy directed forward into the "mouth" of the canyon.


We note for completeness that the last few revolutions of the moonlet around Mars in the atmosphere prior to impact would probably cause very high winds to be generated in the atmosphere prior to first contact. This might cause a planet-wide dust storm of unprecedented proportions, which would tend to fill in and smooth out the canyon bottoms from massive dust deposition. This obvious corollary of the hypothesis suggests the intriguing notion that regular Martian dust storms, whose origin is poorly understood, may be excited by large meteoroid impacts, or by "tidal waves" sent through the Martian atmosphere by such meteoroids.


The steep sides of the VM canyons must plainly show rhythmic sculpting patterns expected to occur when an irregular-shaped moonlet rolls rapidly across the surface. Such odd patterns are present, and are a natural and inevitable consequence of the decayed-moonlet hypothesis, but must be explained in an ad hoc way by the Mars-quake hypothesis. See Figure 3.


In summary, the following elements fit into a complete picture consistent with the hypothesis that Valles Marineris on Mars is a canyon formed by the grazing impact of a population of former moonlets of Mars:


*-There has been an excess of objects striking the Martian surface at angles of less than 15 degrees. Such an excess implies a former population of orbiting objects.

*-VM lies near the Martian equator.

*-VM is oriented parallel to the equator.

*-VM was formed from west to east; that is, in a direction prograde with respect to the planet's rotation.

*-The highest surface features in the equatorial region of Mars are the Tharsis volcanoes to the immediate west of VM.

*-Multiple parallel troughs suggest multiple moonlet impacts, or multiple fragments from one moonlet.

*-The sculpted or "wrinkle ridge" appearance inside the canyons is consistent with the roll of an irregular-shaped asteroid.

*-Phobos will follow precisely this scenario when it decays onto Mars in 40 million years.