Determining the age of surfaces on Mars

Mike Caplinger, Malin Space Science Systems February 1994

A fundamental problem in planetary science is determining how the surface of a planet has changed over time. This tells us something about the dynamics of planetary interiors (for example, how often volcanoes erupt or how often earthquakes occur) and also something about the processes that affect planets from without (for example, how likely it is that a giant asteroid might hit the Earth and change it radically.)

The simplest way of describing how a surface has changed over time is to describe the age of each part of that surface. For example, one could make a map of the surface, color-coding it such that each color represents a different range of ages. But how are we to determine the age of a surface? On the Earth, we have easy access to the surface, and often, to many other older surfaces buried beneath it. (Obviously, if one surface has buried another, the deeper surface must be older.) In the Grand Canyon, for example, one can see the various layers of rock that have been excavated on the canyon walls, and see the evolution of the surface in the area directly. The process of dating surfaces by looking at the relationships between them is called stratigraphy.

In addition, with the development of radiometric dating techniques, we can determine when a rock was formed or changed state directly, by measuring the amount of materials produced by radioactive decay within the rock. This powerful technique has allowed us to determine the absolute ages of all of the surfaces on the Earth, and those surfaces on the Moon from which samples have been returned by the Apollo and Soviet Luna missions.

For the other planets, like Mars, we will be unable to apply radiometric dating until we can study rocks from their surfaces in a laboratory. The only tools we can use to explore Mars today are photographs taken from orbit. Since the first feature of those photographs one sees are the very large numbers of craters, one asks "what use can we make of craters to determine something about the Martian surface?"

In general, the more craters appear on a surface, the older that surface is. But like most principles in the real world, that rule must be applied with caution.

Our best theory about how the planets formed is that they were accreted from smaller bodies, which kept impacting and adding onto the mass of each planet. Eventually, most of these smaller bodies had hit the planets, and so the rate of cratering tailed off to almost zero. The largest bodies (the ones that would form the largest craters) were used up before the smaller ones, since there were fewer of the larger ones to start with. So as a rule of thumb, the larger a crater is, the older it probably is.

We can roughly divide the history of crater formation into three periods, from oldest to newest:

  1. large and small craters formed
  2. small craters only formed
  3. very few craters formed
Since the planets were formed by the impacts of objects, we would expect all of the planets to be uniformly covered with both large and small craters. However, if we take a map of Mars and plot every crater of diameter 100 km or greater, we get this:

So craters are not uniformly distributed on Mars; instead, there are a few areas with significant numbers of very large craters (greater than 300 km in diameter), most of the rest of the southern highlands have only smaller craters, and all of the northern lowlands have very few craters.

We can color-code these regions, using red for the most heavily-cratered areas (the areas with the largest craters), green for the intermediate areas, and blue for the least-cratered areas.

Very roughly speaking, this is a map of the ages of the surfaces on Mars. The red surfaces were formed in period 1, the green surfaces were formed in period 2, and the blue surfaces were formed in period 3. These three periods correspond roughly to the three martian periods Noachian, Hesperian, and Amazonian (named after regions that approximate those ages.)

If nothing occurred to change the surface of Mars through time, it should all look like a Noachian surface. What happened? Since no one can think of a process that would erase only large craters, there must be something that erases all craters and "resets" the surface to smoothness.

This process was probably volcanic in nature: through time, lava flows buried all of the craters in some areas. Other processes like subsidence or erosion are also possibilities, but these would have had to work differently in some areas than others, and this is not likely -- on the Earth, these forces have destroyed craters everywhere simultaneously. Unlike volcanic processes, erosional processes usually happen everywhere at once.

The area with the very largest craters must be the oldest, since it survives from the Noachian. The area with few very large craters must have had very large craters on it, but some process must have erased them. Since that surface still has many smaller craters on it, this resurfacing must have occurred before the Hesperian ended.

The area in the north with no craters on it must have be resurfaced after the Hesperian ended, since otherwise many craters would have been formed on it.

So, in the Noachian Period, Mars was uniformly covered with both large and small craters. During the Hesperian Period, the Noachian surface accumulated more small craters and the large craters remained, but the Hesperian surface was resurfaced and then subsequently accumulated only small craters. We know that something covered the Amazonian surface after the end of the Hesperian Period, since if that surface had been covered before the Hesperian ended, it would have many small craters on it. But there is no way of knowing when this happened more precisely; we are unable to tell what the Amazonian surface looked like before the end of the Hesperian Period.

The impactors that formed the large basins Isidis, Hellas, and Argyre were so large that there were *never* enough of them to uniformly cover the planet, even in the Noachian. Examine this detailed map of Isidis.

Note that it is very smooth, and is about half surrounded by cratered terrain, and half surrounded by a smooth plain. This suggests that whatever filled in the Amazonian surface flowed into Isidis, but could not cover the southern walls of the basin. Remember from the discussion of the crustal dichotomy that the northern lowlands are lower, and hence easier to cover.

Of course, this discussion is very simplified. There are stratigraphic relationships (like the boundary of Isidis) that are used to determine relative ages, in addition to crater counts. The maps shown here display craters larger than about 100 km, and this can be done more accurately using many thousands of smaller craters (recall from our discussion of crater forms that a 100 km crater is a large crater on Mars.) Also, there are complex curves for the number of impacts of a particular size: the division into three periods is fairly arbitrary.

Now note the locations of the major martian volcanoes on the map.

The major volcanoes can be found on the youngest, Amazonian surfaces. This is to be expected, since the Amazonian surfaces were recently covered by lava and are thus volcanically active. However, it is a misconception to believe that the lava that covered these surfaces came from the visible volcanoes. Instead, these areas are volcanic plains, usually formed from fissures which are then subsequently covered by the same lava that flooded the surrounding areas.

Mike Caplinger (