Mars investigators have speculated that during the early eons of martian time, when the atmosphere was possibly more abundant (thicker, with greater surface pressure), water released by impacts and other processes could be distributed as rainfall. Some think that shallow lakes filled Hellas, Argyre and other large craters for a time.
In some of the above images, and several on pages 19-13a and 19-13b, features that could be described as mountains are displayed. Of course, volcanoes fall broadly into that category. Rims around large craters also are mountainlike. Here is a series of mostly parallel ridgelike prominences that are considered low mountains, found here in a region called Tithonium Chasma:
So, once again we see a planetary body with a great variety of landforms, many caused or affected by impact processes. Some of these are unusual (exotic) including those which may reveal water erosion.
By now, one should be convinced that Mars is a geomorphologist's Paradise. As with the Moon in the earlier days of exploration, landform identification, with educated "guesses" as to modes of formation, has been the prime approach to mapping and interpreting the martian surface. Mars exhibits a great variety of terrains and landforms types. Most are given terms that have a Latin derivation. An excellent summary with numerous examples of these types is found at The Atlas of Mars web site. Click on the terms in the left column which brings up usually many images each displayed by clicking on its entry phrase. It is well worth your time to spend an hour or so looking at the wide range of landforms recorded at this site.
Some of the big surprises were infrequent but distinctive sinuous channels, whose morphology is much more similar to river channels that lava channels. One interpretation holds this morphology as evidence of widespread water in the past, in lakes, groundwater or possibly oceans. Expulsion of copious water initiated some sort of hydrologic cycle involving rain storms and runoff. Most of this water has since evaporated into space, although possibly significant quantities may remain frozen as underground ice. Nevertheless, major water activity has recurred as evidenced by the types of dendritic channeling shown in these Viking Orbiter images:
The region depicted in the top image covers the Juvenae Chasma and Vedra Vallis. These are runoff channels, a type confined to the ancient landscapes. Stream flow is the favored origin, based on comparing them with terrestrial counterparts. Some channels seem to originate at craters, which could imply that subterranean sources released either water or lava, following impact offloading. The drainage pattern in the bottom image resembles terrestrial patterns found in soft sediments or wind deposits.
Collapsed lava tubes, which look like some stream channels, have been found in association with martian volcanoes. Here is one example:
Even more striking is this group of lava tubes on basaltic terrain in Pavonis Mons; seen from above this would look like multiple stream channels:
This Viking image shows a channel called Nirgal Valles that looks much like the sinuous rilles described on the lunar surface. Whether this was caused by lava tube collapse or by fluvial action is not obvious at this scale:
Nevertheless, the resemblance of many of the martian channels to fluvial channels on Earth is particularly evident in the next (MOC) image. Located within the large Newton crater, the dark (windblown sand-filled?), flat-bottomed channels look like some headwater types for streams found on Earth:
The "Jury is Out" on the exact origin of the narrow channels in this MGS MOC image. What can be said is that over much of the depression light-colored wind deposits have been trapped and shaped into dunes resembling large ripples:
A distinctive type of drainage called outflow channelling is typically broad and deep, creating canyon-like depressions. A typical example, seen below, is Ma'adam Valles, some 300 km (185 miles) long, which ends in Gusev Crater (far upper right; see page 19-13a):
This next type of landform (left image) may have been associated with catastrophic scouring during abrupt flooding. A similar example on Earth is the Channeled Scablands of central Washington State in the U.S. that developed in just a few weeks from rapid emptying of a huge dammed lake after a natural breakup. Another indication of strong fluid action is a teardrop-shaped landform in Elysium Planitia (right image) a prime example of shaping by streamlining (analogous to aerodynamic sculpturing), in which water flowing from bottom to top has eroded plains material around the rim of a large crater and has terraced and perhaps redeposited debris towards the pointed end.
If riverlike channels did once carry water over the martian surface, one landform they should produce is a fan deposit made up of the debris carried by the streams until such streams are slowed such as to cause their sediment loads to be dropped. This fan is located in Prometheus Terra:
A prominent distributary fan has now been found in the Eberswalde crater in the southern hemisphere. The delta-like fan is 13 km (8 miles) by 11 km (7 miles) in dimension. Here it is in a Mars Global Surveyor MOC image:
MOC close-ups of parts of this fan show several anomalies attributable to fan morphology. In the first image below, one and perhaps two flat-topped ridges emerge above sculpted out surfaces (exposing layers). These ridges may be made of channel deposits that were more resistant than surrounding deposits so that after general erosion of the fan, these remain as topographic highs:
This image indicates the fan's ridges are made up of rock that is consolidated and hardened:
The next two paragraphs are part of a Science release made by JPL/USGS:
The Eberswalde delta provides the first clear, "smoking gun" evidence that some valleys on Mars experienced persistent flow of a liquid with the physical properties of water over an extended period of time, as do rivers on Earth. In addition, because the delta today is lithified -- that is, hardened to form rock -- it provided the first unambiguous evidence that some martian sedimentary rocks were deposited in a liquid (presumably, water) environment. The presence of meandering channels, a cut-off meander, and crisscrossing channels at different elevations (one above the other), provided the clear geologic evidence for these interpretations.
After the sediments were deposited to form the delta, the material was further buried by other materials -- probably sediments -- no longer present. The entire package of buried material became cemented and hardened to form rock. Later, erosive processes such as wind stripped away the overlying rock, re-exposing the delta. Now preserved essentially as a fossil, the former floors of channels in the delta became inverted, to form ridges, by erosion. Channels can be inverted by erosion on both Earth and Mars. Usually this happens when the channel floor, or the material filling the channel, is harder to erode than the surrounding material into which the channel was cut. In some cases, the channels on Earth and Mars have been filled by lava to make them more resistant to erosion. In the case of Eberswalde, there are no lava flows; instead, the channel floors may have been rendered resistant to erosion either by being better-cemented than the surrounding material, or composed of coarser-grained sediment (such as sand and gravel as opposed to silt), or both.
The consensus as of 2007 is that many - perhaps most - martian channels were carved by running water at times in martian history when liquid water was much freer to flow in copious amounts over the surface; probably the martian atmosphere was denser in those times. A plot of major channels in the non-polar regions of Mars reveals an interesting pattern:
What seems mysterious about this pattern is that most of the channel systems end in a "blank" (black) part of the map. Thus the majority of systems were independent and did not connect with each other. The presumption is that each channel emptied into a relatively small region of Mars. In such a region the water would connect with a standing body of lake-sized proportions. While this in itself does not rule out oceans, those would have existed before channel-cutting. The question that emerges when trying to explain the widespread distribution of sedimentlike layers over much of Mars is whether these were dominantly lake deposits or could at least some represent a more continuous marine stage.
One thing is now sure. Some process(es) is/are still producing new channels. This image below shows the same smooth (sandy?) plains that ends abruptly in a cliff. On the left, the image indicates no channeling; on the right at a later time, a well-defined small channel network has since formed. What caused this remains conjectural; wind erosion is one suggestion.
Found both on Earth and on Mars, the "inverted channel" results when a normal channel is filled with sediments that become well-cemented, and later erosion removes surrounding, softer deposits, leaving the channel to stand above the general surface (indicated by shadowing). This is a good example:
Mapping, using largely MGS imagery, has displayed the distribution of channels on Mars. The map below is taken from this website that summarizes the current thinking on martian channels. The type called outflow channels is shown in red and valley networks in yellow.
A map - shown below - quantifying the density of channels on Mars was made public in November of 2009:
The greatest number of channels occur in the higher terrain in the central to southern part of Mars, that is, in the Highlands. Note that this band of channels is not far from the vast blue areas that denote the northern Lowlands. Many Mars scientists believe this blue area once held an ancient ocean. Water from this ocean that evaporates would produce rainfall that tended to land on the more northern Highlands and would be much less towards the South Pole. This rain would collect as runoff in streams that produce the channels.
Another landform that, on Earth, is almost exclusively formed by water-involved erosion is the mesa (or its smaller form, the butte); on Mars it is termed "mensa". This terrestrial landform occurs usually when a more resistant layer series is on top of a weaker, more erodable set. Water, usually within migrating streams, attacks the lower layers, gradually exposing the surface, and causing the upper layers to diminish in size and extent as undercutting erodes into those layers. A residual series of higher landforms results as remnants of the original upper layers on the stripped surface. Here are two excellent examples of a Mars mesa, the first found in Granicus Valles, the second is called Lunae Mensa:
Light-topped mesas are found near Valles Marineris:
Seen closer-up, mesalike blocks of this whitish material shows faint but distinguishable layering. Such units appear to be erosional outliers or remnants of a once continuous sequence of deposits - probably formed by some sedimentary process - found over various regions of Mars.
Well-developed mesas are found at Cerebrus Palin, as shown in this MRO image:
Other variants of the mesa landform include remnants of a thick dark unit found above the lake beds in Aram Chaos (see page 19-13a) and flat-topped CO2 ice "mesas" separated by flat pits in the south Polar ice sheet.