Join Donate

Scientists Search for Distant Life in the Moonlight

Galileo over Jupiter's moon Io


Galileo over Jupiter's moon Io
Galileo arrived at Jupiter on December 7, 1995, and plunged into the planet on September 21, 2003. The image shows the spacecraft's high gain antenna fully deployed, whereas in fact it never fully opened.

On December 8, 1990, the spacecraft Galileo swept by the Earth, picking up a gravity assist for its long journey to Jupiter. As it came within 960 kilometers of its home planet Galileo was wide awake, all its instruments active, taking measurements and collecting data. For the first time Earth was viewed and measured just like any other planet in the solar system, from the perspective of a spacecraft fly-by.

What could a spacecraft learn about our planet in such a brief visit? Would it detect the things we consider most significant about it – its rich vegetation, teaming life, and human presence? When in 1993 Carl Sagan, co-founder of The Planetary Society, published an article with colleagues in Science magazine about Galileo's Earth encounter, these were precisely the questions he was trying to address. To their satisfaction, Sagan and his collaborators found that the evidence for the presence of life, and even intelligent life, was plentiful in Galileo's data. No probe passing by Earth would miss the fact that here was a planet worthy of a more sustained investigation.

The signs of Earthly life picked up by Galileo were numerous: the combination of abundant  water on the surface with unusual amounts of oxygen, methane and ozone in the atmosphere, radio signals, and more. One particular indicator that caught Sagan’s attention was the strong reflection of the near-infrared color in the Earth’s spectrum. This "red edge," he wrote, pointed to the presence of “a light-harvesting pigment in a photosynthetic system.” It was, simply put, the signature of the green plant life that covers vast swathes of Earth’s surface, and it was strongest in those regions of the Earth that are covered with dense vegetation. A probe like Galileo, flying by an unknown planet, would be sure to detect this unmistakable indicator of rich plant life.

The time will probably come when, Star Trek-like, manmade probes will swoop by distant planets and collect data just as Galileo did for Earth. In the present, however, with our limited means and technology, no such close-range encounters are possible. For the foreseeable future, if life-bearing planets are to be found anywhere in the galaxy, they will have to be detected from distances measured by light years, not kilometers. Under such conditions, will the spectral indicator proposed by Sagan and his colleagues be of any use? Will we, in other words, be able to detect that “red edge” in the spectrum of a distant planet, and deduce from it the presence of plant life?

The question is of enormous interest to contemporary astronomers and planetary scientists. So far, admittedly, no Earth-sized planets have as yet been discovered, not to mention imaged, outside our solar system. But the day is undoubtedly near when the sensitivity of the extrasolar planet search will increase to the point when such planets will be found. A new generation of space missions, including the Space Interferometry Mission (SIM), the Terrestrial Planet Finder (TPF), and Darwin, are designed specifically for that purpose – to detect and observe Earth-like extrasolar planets. Within the coming years scientists will be in possession of spectrum measurements, and perhaps images, of distant planets with a mass and orbit comparable to Earth.

Pilar Montañés-Rodriguez, Big Bear Solar Observatory

Pilar Montañés-Rodriguez

Pilar Montañés-Rodriguez, Big Bear Solar Observatory

When this data becomes available, scientists will immediately begin mining it for signs of distant life. The most obvious and easy to detect signature of life, explains astronomer Pilar Montañés-Rodriguez of the Big Bear Solar Observatory (BBSO) in California, is the presence of large amounts of gases in a planet’s atmosphere than can coexist together only because life is sustaining them. Oxygen and methane, or oxygen, water, ozone and carbon dioxide, are examples of such combinations.

But this in itself only points to the presence of simple microbial life, such as existed on Earth for billions of years before the emergence of multicellular organisms. Detecting complex life, such as plants, is much more difficult, she explains, and for that scientists will need to rely on subtle indicators such as the “red edge.” How should we look for these signs of complex life on distant planets, and how likely are we to find them? These are the questions that Montañés-Rodriguez and her BBSO colleague, Enric Pallé, set out to answer.

In the Light of the Moon

A scientist seeking to investigate the spectral signature of a life-bearing planet does not have many options to choose from. Just like Sagan in 1993, Montañés-Rodriguez and Pallé focused their investigations on the only such planet that is known to date – our very own Earth. Since no planet is as well documented and researched as Earth, this would seem at first glance to make things relatively easy. After all, measuring Earth’s spectrum from space is precisely what Galileo did in 1990, and there are satellites in orbit around our planet that are capable of making similar measurements. Surprisingly, however, data collected by spacecraft and satellites simply will not do, explained Montañés-Rodriguez.

Long distance measurements of extrasolar planets, she said, are very different from close range observations such as the one obtained by Galileo. The spacecraft, passing less than 1000 kilometers above Earth’s surface, had a clear view of our planet’s geographical features. When it pointed its spectrometer at regions of Earth where plants are plentiful, such as the Amazon forest, it recorded a strong “red edge” in the spectrum. But when it focused on the ocean, or the Sahara desert, the “red edge” disappeared. The same is true of the satellites orbiting Earth, which can easily focus on different regions at different times, and record their specific spectra.

When observing an extrasolar planet light years away, however, such geographical distinctions are not possible. We will not know, for such a planet, where its oceans and continents lie, what regions are covered with greenery and which are deserts. In fact, we won’t even know if the planet HAS oceans and continents, not to mention deserts and jungles. All we will have is the composite spectrum obtained from the planet as a whole, including possible jungles and deserts, as well as oceans and clouds. In such an integrated spectrum, the “red edge,” originating exclusively from green regions that are not covered by clouds at the time of measurement, would be much diluted. Whether it would be detectable at all is one of the main questions Montañés-Rodriguez and her collaborators set out to answer.


JPL / Greg Redfern


If spacecraft and satellite images cannot provide us with an integrated spectrum of Earth as a whole, where can we obtain measurements of Earth comparable to those of extrasolar planets? The surprising answer, according to Montañés-Rodriguez, is from earthshine.

All of us have known earthshine our entire lives. When we step outside at night time, and look up at the Moon when it is not full, we see that it is divided into a bright and a dark region. The bright region, of course, is a segment of that part of the Moon that is illuminated directly by the Sun. The rest of the Moon appears very dark by comparison, but it does not melt completely into the dark background of space. It has a faint ghostly glow to it, which is the light reflected at the Moon from the Earth. That faint glow is earthshine.

Unlike direct observations of Earth, which measure light reflected from particular regions, earthshine is an amalgam of light from up to one half of the Earth’s surface. Since it possesses a single spectrum combining light from different regions of Earth, earthshine is an excellent representation of how our planet would appear to observers light-years away. By carefully observing earthshine and recording its spectrum, Montañés-Rodriguez and her collaborators sought to discover whether the telltale “red edge,” the signature of extensive vegetation, is detectable in the light reflected from Earth.

The Hunt for the “Red Edge”

Initial results were not promising, as Montañés-Rodriguez and Pallé explain in an article published in the November 1, 2006 issue of The Astrophysical Journal. Using the 60 inch telescope at the Mount Palomar Observatory and a high precision spectrometer, the authors observed and recorded the earthshine spectrum throughout the night of November 19, 2003. Unfortunately, they note in the article, “our analysis of earthshine data . . . did not show a significant vegetation signal in the Earth’s globally integrated spectrum.” The reason for this, they concluded, was cloud coverage. When a significant part of the regions on Earth that contribute to earthshine are covered with clouds, the strong reflection from the clouds drowns out the signal from the green regions. If, as usually happens, some of these regions themselves are covered with clouds, the “red edge” signal is even weaker.

Earth regions contributing to Earthshine at different times on the night of November 19, 2003.

Pilar Montañés-Rodriguez / The Astrophysical Journal

Earth regions contributing to Earthshine at different times on the night of November 19, 2003.

To confirm this hypothesis Montañés-Rodriguez and Pallé decided to compare their results to the actual cloud coverage and green terrain that contributed to the earthshine on the night of their observations. Relying on the detailed records of the International Satellite Cloud Climatology Project (ISCCP) for November 19, 2003, they created a computer model that combined the light reflected from the different parts of the Earth that contributed to earthshine into a single spectrum. When they compared their predictions to the actual observations, the fit was excellent: the computer model accurately predicted the earthshine spectrum throughout the night, as the Earth turned on its axis and different regions were exposed to the Sun.

The superb fit between the computer model and the observations gave the authors confidence that they could now predict the earthshine spectrum for any night of the year, provided that detailed cloud cover information was available. With this powerful tool in hand, Montañés-Rodriguez and her colleagues could now approach the problem from a different angle. Given that the “red edge” in the earthshine spectrum was hardly detectable on the night of November 19, 2003, that does not necessarily mean that it is useless as a means of detecting vegetation on a distant planet. Perhaps there are other times when the signal is prominent, and if so, on which nights should we expect to detect it?

Our first reaction, at this point, might be that we should look for the spectral signature of vegetation on clear nights, with few clouds. A moment’s reflection, however, will reveal our mistake. Earthshine, we need remember, is not derived from our particular location, where clouds may or may not be present, but is a reflection of a large segment of the Earth. On such a broad scale cloud coverage never changes much, and always extends over about 60% of the surface of the planet. Earth, in other words, will be about equally cloudy no matter what day we choose to conduct our observations. On which days, then, should we expect to find the strongest ‘red edge” signal?

Three days of earthshine

Pilar Montañés-Rodriguez / The Astrophysical Journal

Three days of earthshine
The percentage of earthshine-contributing regions of Earth that are simultaneously cloud-free and covered with vegetation, on three different days. The blue line represents the observation data for November 19, 2003, when close to 40% of Earth's surface contributed to earthshine. The black and red lines represent December 19, 2002, and December 7, 2003, when only a narrow sliver of Earth contributed to earthshine. The three peaks on these dates represent, from left to right, Asia, Africa, and South America.

To find out, Montañés-Rodriguez and Pallé decided to model the earthshine spectrum for every day of the year 2003. Unlike the observation data for November 19, 2003, a few of these days, such as December 19, 2002, and December 7, 2003, showed sharp fluctuations in the “red edge” signal over the course of 24 hours. The signal would appear, strong and clear and then disappear, and so repeatedly up to three times in succession. When the authors examined the data for those dates, they found that the three peaks corresponded to times when the regions contributing most to the earthshine included the plant-rich regions of Asia, Africa, and South America in succession. On these dates at least the earthshine spectrum strongly suggested the presence of plant life on Earth.

In order to understand why the “red edge” signal, so weak on the original night of observations, was clearly detectable on the additional modeled dates, we need to consider that the size of the area of the Earth that contributes to earthshine varies enormously. Just as the Moon waxes and wanes when seen from the Earth, depending on which part of it is lit by the Sun, so the Earth waxes and wanes every month when viewed from the Moon. Naturally, only the brightly lit areas of the surface of the Earth which are seen from the Moon contribute to earthshine. These regions vary from a tiny sliver when the Earth is “new,” to a bright disk covering half the planet’s surface when the Earth is “full.” The days when models showed the “red edge” in the earthshine to be particularly distinct, were also days when the Earth was almost “new.” This means that if viewed from the Moon, Earth would appear as a thin sliver showing a small fraction of the planet’s surface.

On most days of the year earthshine is composed of light from a relatively large segment of the Earth’s surface. Inevitably, a large proportion of this area is covered with clouds, and it is likely to be composed of both land and water regions. Even as the Earth turns on its axis, reflecting light from different continents and oceans towards the Moon, the overall effect does not change much. At all times, the “red edge” signal, coming from unclouded land regions with high vegetation, is drowned by the light reflected by clouds or the ocean. That, indeed, proved to be the case on November 19, 2003, when Montañés-Rodriguez and Pallé conducted their observations.

The spectrum of earthshine and the 'red edge'

Pilar Montañés-Rodriguez / The Astrophysical Journal

The spectrum of earthshine and the "red edge"
The black line represents the spectrum measured on November 19, 2006. The green line represents a computer modeling of the spectrum for the same date, showing excellent agreement with the observed spectrum. The red line represents the modeled spectral albedo for December 7, 2003 at 21:00 Universal time. The "red edge" is represented by the slope of the line between 680 nm and 740 nm. Note that the slope for December 7, marked by the broken line, is much steeper than that for November 19.

But when only a tiny sliver of the Earth contributes to earthshine, the case is different. As the Earth rotates, this sliver will be composed of oceans and continents in succession. Because it is so narrow, however, it is unlikely to contain large swathes of both at the same time. When it is over the Americas, for example, a large percentage of the light reflected towards the Moon will indeed originate from the American landmass rather than from the Atlantic and Pacific Oceans. This means that when earthshine is composed mostly of light from high vegetation areas, its “red edge” signal will be easily detectable, and will not be swamped by light coming from oceans and deserts.

The same applies for the problem of cloud coverage. On most days, areas covered by clouds contribute a large and more or less fixed proportion of light to the earthshine. But when the contributing segment of the Earth is composed only of a thin sliver, things are different. At times, no doubt, the sliver will be completely covered with clouds, and no “red edge” will be detectable in the earthshine. At other times, however, the thin sliver will pass over areas that are clear of clouds, and if those coincide with high vegetation regions then the spectral signature of plant-life will be easily detectable.

To put it differently, when the Earth is “new” the thin crescent of light that contributes to earthshine moves like a scanner beam over the surface of our planet. As the Earth rotates, and the “beam” passes over green landmasses, its spectrum will reveal a strong “red edge” signal at least some of the time. Conversely, when it passes over oceans, no such signal will be discernable. This is sufficient to provide the rough topographical resolution evident in the earthshine of December 19, 2002, and December 7, 2003: when the “scanner beam” passed over the green landmasses of east Asia, Africa, and South America, a “red edge” was clearly in evidence; in between, the signal disappeared in the vast stretches of ocean and desert.

The conclusion for Montañés-Rodriguez and her colleagues was clear. Earthshine in general is a poor indicator of the abundance of plant life on our planet. On most days the telltale “red edge” in the spectrum, which points to vast stretches of greenery on the Earth’s surface, is drowned out by the signal from oceans and arid land regions. But on those days when only a narrow band on the Earth’s surface contributes to earthshine, the “red edge” shows up clearly in the spectrum, rising and falling with the geography of the sunlit sliver.

Earthshine-contributing areas on November 19, 2003, and December 7, 2003

Pilar Montañés-Rodriguez / The Astrophysical Journal

Earthshine-contributing areas on November 19, 2003, and December 7, 2003
On November 19 (left) most of Earth's disk appeared lit when seen from the Moon, and therefore contributed to earthshine, whereas on December 7 (right) only a sliver was lit. Note that on November 19, only about 15% of the lit area was covered with vegetation. On December 7, when only a sliver was shining, a full 48% of the Sunlit areas were covered with vegetation. Cloud coverage has been omitted from these images.

From Earthshine to Other “Earths”

What does all this teach us about detecting life on extrasolar planets? At first sight, it seems, not much. Earthshine, after all, is determined by the unique geometry of the Earth – Moon system, and its relation to the Sun. When future missions will attempt to measure the light from distant planets, they will surely look for light reflecting directly off a planet, not light bounced off a hypothetical moon. Indeed, separating the light from a distant planet from that of its star is enormously difficult in itself; detecting light from the planet that is, in turn, bounced off a moon, is quite impossible. The case of earthshine would appear to be unique.

Montañés-Rodriguez and her fellow authors, however, suggest otherwise. Extrasolar planets, they explain, do have “phases” when viewed from a great distance, which are very similar to Earth’s “phases” when observed from the Moon. When a planet is on the far side of its star, as viewed from Earth, it will appear to be “full,” presenting us with a shiny bright disk covering half the planet’s surface. As the planet orbits its star, it eventually arrives at a point where it is closest to us, positioned in between Earth and its star. At that point the planet disappears from our view, since it reflects all the light back towards its star. But just before and just after the planet arrives in this position, when it is “new,” it presents us with a narrow sliver of light. Just like regions of Earth contributing to earthshine, the light from this sliver represents a narrow band on the planet’s surface; and just as in the case of earthshine, the band “scans” the surface of the planet as it rotates around its axis. It is on those days, just before and after the planet comes between us and its star that we are likely to discern a “red edge” signal – the signature of plant life on a distant world.

The similarities of observing the Moon and extrasolar planets

Pilar Montañés-Rodriguez / The Astrophysical Journal

The similarities of observing the Moon and extrasolar planets
Just as the Moon waxes and wanes, so does an extrasolar planet when viewed from Earth, as long as the line of sight is close to the planet's orbital plane. The best time to look for the "red edge" signature is when the planet is "new," and only a sliver of it appears lit by its star.

According to Montañés-Rodriguez, the implications of her study for the search for life on distant planets is a mixture of good and bad news. The bad news is that in order to detect a “red edge” in the spectrum of a distant planet, we need to be observing it along a line close to the plane of its orbit. Planets whose orbital plane deviates greatly from our line of sight will not exhibit the “phases” that make the detection possible. Furthermore, even if the geometry is suitable, and we are observing the planet from the right direction, the fact remains that the “red edge” will be detectable only at the times when the planet itself is faintest – i.e. when only a small sliver of it is brightly lit. As it happens, those are also the times when the planet is particularly close to the star, as seen from Earth, since it is just before or after it passes right in front of the star. The combination of a very faint planet, positioned very close to its star, will make it exceedingly difficult to isolate the light from the planet from that coming directly from the star. As a result, Montañés-Rodriguez estimates that the hunt for the “red edge” in the spectrum of distant planets will require instruments ten times more sensitive than those on any currently planned mission.

The good news is that the red edge is there in the planets’ spectrum, and ultimately detectable when we know when and where to look. Even if the current generation of missions is not sensitive enough to take advantage of this, other missions, ever more advanced and sensitive, will undoubtedly follow. And when they do we just might find that our Earth is not alone but a member of a large galactic family of planets, all of them green, rich in water, and teeming with life.

Bill Nye and people
Let's Change the World

Become a member of The Planetary Society and together we will create the future of space exploration.

Join Today

The Planetary Fund

Help advance robotic and human space exploration, defend our planet, and search for life.


You are here:
, , , , , ,