With the discovery of some of the youngest (~ 100 million years) volcanic deposits yet recognized, a new study suggests that volcanism on the Moon may have continued for an extended period of time – ending more recently than once believed. One hundred million years sounds very old, and it is by terrestrial (Earth) standards of geological age. But we’d thought that lunar volcanism was largely finished around 3 billion years ago (with some minor additional eruptions occurring as “recently” as about 1 billion years ago). If confirmed, this result could mean that our understanding of lunar thermal history is wrong and will need to be revised.
Using just pictures (admittedly very high-quality ones), how do we “date” unvisited places on the Moon and planets and determine their origins? And what exactly do we mean when we give a “date” for some geological event, one that clearly occurred long before humanity existed?
There are two different concepts that geologists use to think about deep time. An age can be either relative or absolute. Relative age is based on the very simple but powerful concept of the geological law of superposition – younger geological units cover or intrude older units. A crater on top of a volcanic plain is younger than (post-dates) that plain. Such a seemingly trivial observation is a very powerful tool used to decipher planetary history. Virtually every geologic map showing the spatial and temporal relations between recognized units is made from spacecraft images using this concept.
Although relative age can reconstruct the sequence of unit deposition, it says nothing about the actual duration of events or the length of time that has elapsed since. To address these issues, geologists determine absolute age – the length of time (usually in millions of years) since some geological event occurred. Absolute ages are determined through the measurement of radioactive isotopes within a rock. Radioactive elements decay at rates known from laboratory measurement. For volcanic rocks, isotopic systems “close” at the time the rock crystallizes, so this measurement technique determines the length of time since the lava flow erupted. To ensure that the analyzed sample has a well-understood history and context, this technique requires both precision laboratory analysis and careful geologic field study.
Long before astronauts landed on the Moon, we’d mapped the stratigraphy (study of layered rocks) of geological units from photographs. Using the powerful technique of superposition described above, the relative ages of units and the global history of the Moon were pretty much nailed down prior to the Apollo landings. However, we did not know the absolute ages of lunar units. Although some fairly good educated guesses were made, we couldn’t be certain if they were even in the ballpark until we acquired and analyzed some actual lunar rock samples. The early Apollo 11 and 12 landing sites were located in dark, smooth maria, which we learned are composed of basaltic volcanic lava flows. From laboratory study of returned samples (plus knowing where in the field they were collected – context), we know that the surface rocks of the Apollo 11 site are about 3.6 billion years old while the lavas from Apollo 12 are about 3.1 billion years old (close to a 500 million year difference).
Counting craters is another technique we use to establish age. Simply put, the longer a surface is exposed to the bombardment of cosmic debris, the more impact craters it accumulates. Thus, older surfaces have a higher density of impact craters than younger ones. Obligingly, the crater density of the Apollo 12 site is indeed lower than that of the Apollo 11 site, confirming this seemingly airtight logic. I say “seemingly” because, although crater counting is simple in concept, complicating issues arise in practice, such as the inadvertent inclusion of non-impact features (volcanic collapse pits) and secondary impact craters formed by debris thrown out of a primary crater – such features can make a surface appear anomalously old. Usually, we can distinguish these features by careful mapping, but not always. However, the Moon’s geologic history is almost “pristinely” preserved compared to the dynamic, constantly evolving surface features of the Earth. Our study of the impact history of the ancient Moon gives us valuable clues about the Earth’s impact history, where much of that early geology has been wiped away.
Subsequent lunar landings and study of their samples helped to “fill-in” the geological time scale for the Moon. Four Apollo and two Russian Luna missions returned samples from specific mare sites. This allowed us to calibrate the relative age scale based on crater density with absolute ages. That information was (and still is) used to estimate the absolute ages of mare deposits all over the Moon and by extrapolation, other distant terrestrial (rocky) planets – Mars, Mercury and Venus – as well. Thus, you will often see absolute ages given for unvisited sites and other lunar geological events, as well as for features on Mars.
So, after establishing this wonderful technique to determining ages, what could possibly go wrong? Plenty could and often does – science is never “settled.” First, the crater density time scale is not perfectly understood, especially at its youngest end. Because of uncertainties in the absolute ages of the most recently dated surfaces on the Moon (the craters Copernicus and Tycho), we cannot be certain whether the youngest large mare lava flows are 1 billion, or over 2 billion years old (a greater than 100% relative error). Second, all of the newly discovered volcanic features in this recent study are very small (only a few hundreds of meters across). The effect of this diminutive size is to produce large error bars in the gathered statistics for crater density; the smaller (fewer) the population, the larger the relative error. Thus, we cannot get crater density numbers as precise as those from older surfaces that cover a larger area. Finally, the new features still have an unknown composition (we have no returned samples from this location). They look like smooth lava flows, but are they? For example, if the smooth deposits consist of some type of fluidized particulate debris (such as an ash flow), they might not retain a cratering record comparable to that preserved by a lava flow.
Counting craters to estimate absolute age is a dull-knife technique. Despite the most rigorous protocols, inadvertent errors can and do creep in. Even when we have samples of a given unit, assumptions must be made that the sample has remained undisturbed since its formation and that it represents the lava flow (unit) as a whole. Any time you construct an edifice of knowledge as a house of cards, it can collapse under the slightest pressure. But when it’s all you have, you use it. And for the most part, crater statistics combined with radiometric ages from returned samples have helped us to better (and more completely) understand the history of the Moon.
We’d be much more certain in our understanding of the sequence of geological events on the Moon and other planets if we had more samples collected from many different units that are widely spaced in time. It is largely for this reason (among others) that sample return is always a high priority on the “wish lists” of missions compiled by planetary scientists. Unfortunately, such missions are very difficult to conduct because you have to get into (and then out of) the large gravity well of a planet, which makes the mission very expensive. Several groups are investigating alternative means of getting this information – less costly concepts that will enable sample return and remote measurement of absolute ages on other planetary surfaces. If these other means are successful, we may yet be able to obtain the critical information needed to better understand the history of the Moon and planets.