There are eight planets in our Solar System. Of the eight planets three of them have silicate, probably active, volcanism. Mars, Earth and Venus. It is very insightful to compare the volcanism between these three planets in order to find out more about how Earth’s own volcanoes work. Yet I find that extra-terrestrial volcanism is still misinterpreted despite the wealth of data that we have. Some popular ideas are still running around which I think could be disproven easily. Particularly the theory of how the Martian outflow channels were formed.
The Mariner 9 was the first spacecraft to orbit Mars. This was in 1971. After a global dust storm cleared away it was finally able to take pictures of the planet surface. The images it took revealed the existence of giant canyons, and river-like features. Valles Marineris, Ares Vallis, Kasei Valles and other landforms which were interpreted as having formed due to flowing liquid water. But it has been a while since then. Nowadays we have a global mosaic of the surface of Mars, at a resolution of 6 meters per pixel, taken by the Context Camera of the Mars Reconnaissance Orbiter. This is incredible detail that allows to look closely at these outflow channels and know more about how they formed.
First I should clarify that there are two different types erosive structures that have been attributed to water. One of them is the valley networks. The valley networks occur mainly in the Southern Highlands of Mars and are usually 4 billion years old. They have features like those of river drainage basins on Earth. To me they look water made and clearly have nothing to do with volcanism, could have been rivers of liquid water. Although the idea of water or carbon dioxide ice glaciers should also be considered.
The second type is outflow channels. These are much younger, some of them only a few millions of years old. They are found close to volcanoes, most of them being concentrated in the Tharsis and Elysium volcanic provinces. They lack sedimentary deposits like deltas or ripples. The channels initiate abruptly from a chasm or depression in the ground which has led to think that they were formed in huge outbursts of groundwater contained in aquifers under the surface. It is also remarkable the size reached by some of the outflow channels. The largest one is Kasei Valles, a canyon with a length of 1600 km, a width of up to 400 km in places, and a depth of up to 3 km.
One of the main problems with the liquid water hypothesis for outflow channels is that Venus has identical valleys to those of Mars. It not possible for liquid water to exist on the surface of Venus. Additionally the youngest canyons of Venus always are seen to channelize lava flows, and they tend to emerge from radial and circumferential fractures around volcanoes.
I’m here to explain why there is morphological evidence, unambiguous evidence I’d say, that the outflow channels of Mars are lava made. Not water but lava. The result of giant flood basalt eruptions. I’m not the first person to say so and for example the geologist David Leverington has done some great articles about this aspect. This post will focus only in the morphological evidence, so for a full discussion on why the outflow channels were made by lava, and also why the groundwater hypothesis has many problems, I will leave one of his articles here:
The formation of fossae and catenae
There is something important which is often overlooked and is that lava erosion may occur not only along lava channels, but also along dike intrusions. Magma flows underground through intrusions which for simplicity I simply refer to as dikes, although there may be other morphologies like sills or cone sheets that are involved. In Hawaii, in particular, there exist many features similar to the so called fossae and catenae of Mars, only much smaller, that are related to dikes and their erosive processes. Volcanism on Mars is of a huge scale so that it does the same landforms only bigger in scale.
In the island of Hawaii, during the eruption of Mauna Ulu, in 1969, the pit crater Alae had been filled with a rootless lava lake. Rootless means that it is not fed from below. The lake formed from lava that erupted out of Mauna Ulu and then flowed downslope towards Alae where it collected passively. On August 4 the lake catastrophically drained into a dike intrusion. The lava lake emptied in half an hour at a flow rate of 5500 m3/s. Spectacularly high for terrestial lava flows although probably insignificant compared to Martian effusions. The lake drained through a dike along which erosion took place, a substantial amount of rock was carried away with the flow leaving a chasm that was 10 meters wide, up to 70 meters deep, and 800 meters long. Similar features are also formed when Nyiragongo volcano in the Congo has its typical catastrophic lava lake outbursts.
What is also interesting about this event is that it was rootless dike intrusion. It was fed from a surface lava lake. Lava flowed back into the ground making a dike that was probably very shallow, just skimming the surface, as shown by the fresh lava that could be seen in the bottom of the chasm. If there had been someone looking into the fracture while the lava lake drained it would have probably been able to see the raging torrent of lava. I think that rootless dike intrusions might be extremely abundant in Mars from what I have seen.
Later in the Mauna Ulu eruption there was another demonstration of the erosive power of lava/magma. Magma flowing along a dike from the summit of Mauna Ulu to Alae gradually ate away the rock and created a series of pit craters which later merged into a continuous trench 40-60 meters wide.
Other landforms commonly seen in the surface of Mars are the catenae. These are chains of pit craters often related to grabens, and from which lava can sometimes be seen to emerge. On Earth structures similar to the catenae of Mars occur along eruptive fissures and seem to be linked to dikes, possibly another form of dike erosion like fossae, but discontinuous. Certain portions of the dike may widen due to erosion so that the roof collapses into the magma, the rock is carried downstream, then the roof continues to collapse incrementally making a pit crater. Such collapse structures are often located at the upslope end of an erupting fissure that lasts long enough for the erosion to show up.
The creation of an outflow channel
Lava is known to be able of eroding into the bedrock it flows over. One of the probably many examples that exist on Earth is the Kazumura lava tube in Hawaii. This is the longest lava tube know on Earth, and formed during overflows from the summit of Kilauea in the 15th century. Speleologists have descended into the Kazumura cave. The lava tube shows meander migration and downcutting into the bedrock, similar to processes that take place in rivers. Some meanders migrated up to 9 meters downstream. Certain parts of the lava tube with a steeper slope were eroded backwards, this back-cutting resulted in the formation of lavafalls, and at the bottom of the cascade the turbulent flow created large pools which expanded laterally through erosion making a wider passage. So yes, lava erosion happens. It is thougth to be due to lava melting the rock that it comes into contact with. On Earth it is a more subtle process and only observed upon close inspection of lava tubes or channels. On the other hand Mars, with its gigantic eruptions that are far more intense, longer, and more voluminous than those on Earth, it can erode spectacular canyons.
The outflow channels have been traditionally attributed to water. Outbursts of water from aquifers. However there are some characteristics of outflow channels that are not consistent with water. First is that there are no sedimentary deposits, there are no deltas, nor ripples, nor any sort of sedimentary structure that is characteristic of flowing water. And second they have features that could have only been done by lava. As an example I will use Olympica Fossae, a 650 kilometres long outflow channel that switches between channel and dike transport. It is located in the volcanic province of Tharsis and it maybe emerges from a fissure of the volcano Alba Mons.
Right below is an image of Olympica Fossae. In this particular spot the canyon is a system of braided channels that reaches 7-8 kilometres wide and 500 meters deep. Here the channel has just emerged out of the large fossa to the right. There are a series of other small fossae visible along the sides of the valley. It doesn’t appear possible that water could have made structures such as these, however lava would. These linear features might be rootless dike intrusions which propagate from the lava channel outwards and have eaten away the rock closest to their source. It can also be appreciated that two smaller channel systems emerge out of these fossae. Lava that was flowing down the main, master channel would have entered the rootless dikes and emerged from fissures located at slightly lower elevations.
Some of the smaller streams from braided channel network can be seen to flow away from the rest and turn into normal looking lava flows. That is where you clearly see that the outflow channels are lava rivers which have been active for so long that they have melted down into the bedrock, farther down however they become normal streams of lava. For example this is a flow that separated from the braided network shown in the previous images, it is no longer erosive and instead has the typical appearance of sheet aa flows like those seen in intense volcanic eruptions:
Right below you can see the big picture. The outflow channel in reality is only a small portion of a much larger flood basalt eruption. The eruption of Olympica Fossae I estimate may have been around 3,000 km3 of basaltic lava, perhaps even more since there is a possibility that Olympica Fossae was just one of multiple erupting fissures in the area. The total length of the flow seems to be 1900 kilometres, but only the upper 800 kilometres have experienced significant erosion
At the start of the eruption lava poured out of multiple fissures making many lava streams. The streams separate and merge again in fascinating patterns and shows that the whole system of channel was active at once. Later activity focused into fewer channels and eventually was limited to the master channel which cuts deeply into the ground. The transition between erosional and non-erosional lava channels is very well visible.
There is one location where the master channel goes into a dike and travels 170 km through a fossa before reappearing again farther south. The master channel and the subsidiary streams seem to have a continuation downslope across the fossa as normal lava rivers. It can be seen that in the early phase of the eruption the main channel started a rootless dike intrusion which then diverted the flow of lava 170 km away, changing the path followed by the lava.
Thus there is obvious evidence that Martian outflows valleys were created by giant floods of lava. So it is time to discard the idea of groundwater outbursts. There is just better information now. When inspected closely it shows the true lava-made nature of outflow channels.
Just to finish here are some images of the youngest outflow channels in Mars, those of Marte Vallis. The last eruption of Mars took place perhaps as recently as 3 million years ago, it erupted from Cerberus Fossae and created the Athabasca and Marte Vallis. Dark material is rugged aa-type lava. Bright material is either smooth pahoehoe-type lava or old Martian rock. If you look closely you can appreciate the broken plates of black aa lava carried by the stream, and where they separated from each other molten material rose up from below making a smoother surface of pahoeheoe. The convoluted braided streams of aa lava which at places merge into a spectacular 100 kilometres wide river of lava, can be seen to have excavated banks against the obstacles it encountered. Erosion is slight but noticeable. This was a much more brief eruption than the one that made Olympica Fossae, however it was superior in intensity and volume.