The Earth in time as a template for the characterization of inhabitable exoplanets

Sanromá Ramos, Esther

2015-A
Descargar PDF

Resumen

The detection of planets beyond our Solar System has increased rapidly in the last few years. Given the rate of discoveries, the interest that this blooming field of science inspires in the astronomical community, and the planned and ongoing instrumental developments, it is very likely that true “Earth twins” will be discovered in large numbers in the near future. The fast pace at which the atmospheric characterization of already discovered exoplanets is progressing has led to some interesting questions, such as: “How will we recognize a habitable or inhabited distant planet?” or “What kind of biosignatures should we expect and look for from such worlds?” The logical starting point to answer these questions is to study the Earth, the only planet that is known to support life. The study of our planet will be essential for understanding future observations of planets similar to our own, and also for the characterization and the search for life elsewhere. Its study will help us to explore those signatures unique to life, or evidences of habitable conditions, that could potentially be detectable from an astronomical distance.

In the last years, several studies both theoretical and observational have been carried out with the aim of determining how our present Earth would look like to an extrasolar observer. In this thesis we extend these studies to time scales of millions of years, throughout the Earth evolution, by studying how different continental distribution, cloudiness levels, atmospheric compositions and/or the evolution of different life forms, could have affected the photometric and spectroscopic characteristics of Earth as a whole.

The results of this thesis will allow us to better understand the signatures of life in our own planet along its history, and to be prepared for the identification of spectroscopic evidences of habitable conditions and life in the future observed habitable/inhabited extrasolar planets.

INTRODUCTION

Although most of the discovered exoplanets so far are gas giants, it has been found that in our galaxy small planets are more common than large ones (e.g., Cassan et al. 2012). In fact, the search for extrasolar planets has already yielded the discovery of some Earth-size and even Moon-size exoplanets, including the discovery of planets in the habitable zone of their stars (e.g., Barclay et al. 2013; Borucki et al. 2013; Gilliland et al. 2013). The pace of discoveries is so fast that the research frontier is shifting from finding exoplanets to characterizing their properties. Absorption lines of elements such as water, methane, carbon dioxide or carbon monoxide in the atmosphere of extrasolar planets have already been reported in the literature (e.g., Tinetti et al. 2007; Swain et al. 2009; Brogi et al. 2012). The detection of this kind of signatures in the atmosphere of small planets similar to Earth is still out of reach, though. This is a very exciting scenario which has made it increasingly possible that within the next decade or so, we will be able to discover potentially habitable or inhabited exoplanets, which we will need to characterize.

A variety of studies have been carried out aiming at determining the possibility of recognizing if a distant world is able to harbor life. One of the approaches has been to investigate the bodies of our own solar system. Several authors have suggested that it might be possible to find habitable environments different from Earth’s within the Solar System. Some of these places are Titan’s hydrocarbon lakes, Enceladu’s subsurface, or the Galilean moons of Jupiter (e.g., Spohn & Schubert 2003; McKay & Smith 2005; Parkinson et al. 2008). But if life has depleted in these environments, it does not leave a trace which is remotely detectable. Moreover, taking the Earth as our sole example of inhabited world, several authors have put their efforts into the characterization of our planet (e.g., Pallé et al. 2003, 2004; Cowan et al. 2009; Robinson et al. 2011). Many of these works have analyzed the Earth’s atmosphere, aiming at retrieving information about the existence of life on it, particularly looking for the presence of biosignature gases (e.g., Des Marais et al. 2002; Turnbull et al. 2006; Sterzik et al. 2012). Other authors have attempted to determine if it would be possible to remotely differentiate surface reflectance signatures related to pigments, such as the red-edge of leafy plants, (e.g., Seager et al. 2005; Montañés-Rodríguez et al. 2006; Kiang et al. 2007a), and the presence of purple bacteria in the early Earth (Sanromá et al. 2014), or to differentiate environments that can support extreme life forms (e.g., Cockell et al. 2009; Hegde & Kaltenegger 2013). Even the possibility of detecting hypothetical exovegetation have been investigated (Tinetti et al. 2006; Kiang et al. 2007b).

However, as extrasolar planets are expected to show any stage of evolution, it is not only interesting to study and characterize the observables of the Earth as it is today, but also at different epochs of its history (e.g., Meadows 2006; Kaltenegger et al. 2007; Sanromá & Pallé 2012; Sanromá et al. 2013, 2014). Since our planet was formed 4.5 Ga ago, it has undergone multiple changes in its atmospheric composition, temperature structure, and continental distribution. Moreover, our planet has been inhabited for about 80% of its life (Mojzsis et al. 1996), and the predominant life forms inhabiting its surface have also changed. Since Earth’s primitive microbes colonize a great diversity of environmental settings and utilize a wide range of chemical and energy resources (e.g., Seager et al. 2012), it is not unreasonable to imagine that other water and carbon-based life use the same resources on other planets, and hence produce similar biosignatures of life. Thus, the Earth’s history provides a suite of environments different to modern Earth, which serve as analogues for habitable exoplanets, making it a perfect test bed to study and describe life’s most relevant fingerprints in order to recognize them in other worlds.

DESCRIPTION OF THE WORK

In this thesis we have presented a comprehensive study about the Earth, the only planet known to support life, at present and over its geological evolution. Here, we have explored both the photometric and spectroscopic characteristics of our own planet as if it were seen from an astronomical distance with the aim of being prepared for the characterization of future detected Earth-like exoplanets. We have studied how different atmospheric composition, continental and cloud distribution, as well as the evolution of different life forms might have affected the way our planet looked from afar. In this section we briefly summarize the three main parts of this work and their results:

1. RECONSTRUCTING THE CLOUD DISTRIBUTION AND LIGHT CURVES OF EARTH ALONG ITS HISTORY

Clouds are one of the most important parameters in the global energy balance of our planet and have profound interactions with weather and climate. In the first part of this thesis we attempted to provide an understanding of the Earth’s large-scale cloudiness behavior using satellite-based estimations. We studied how clouds distribute themselves over the Earth’s surface, and classified them according to the underlying surface and the geographic latitude obtaining a semi-empirical cloud model.

We found that our cloud model reproduces well the general features of Earth at present. Especially the differences in cloud cover between land and water at the same latitude, and the latitudinal range 40º- 75º, both North and South. However, there are some small-structures mainly over oceans that our model is not able to reproduce. These differences are due to oceanic effects, such as sea surface temperature, trade winds, ocean currents, and regional meteorological phenomena such as «El niño»/»La niña», which can influence cloud formation considerably. This cloud model was then used in an attempt to reconstruct the possible cloud distribution of past historical epochs of Earth, such as 90, 230, 340 and 500 Ma ago, when the landmass distributions were very different from today’s.

As Bond albedo is one of the parameters that control the Earth’s temperature and hence its climate, one of the possible mechanisms that have been proposed to explain how Earth was kept warm during the Achaean, even though the Sun was 30% dimmer than it is today, is the development of continents and their distribution. Moreover, future detected Earth-like planets are expected to show any stage of evolution and any continental surface distribution would be possible. Hence, in the first part of this thesis we also explored the effect that considering different continental configurations has on the visible light reflected by Earth and the changes that these different distributions cause in its daily light curves. To do that, we used a simple albedo model and the reconstructed cloud distribution of the different continental distributions considered.

With the results obtained in this part of the thesis, we concluded that the mean albedo value of Earth has remained approximately constant along its recent history, around 0.3. However, daily variations in the past may have been larger than present day’s. In particular, daily variation of Earth 500 Ma ago was three times larger than present’s day. This increased variability could help in the determination of the rotational period of the planet seen from an astronomical distance. Moreover, these higher mean albedo values should have had profound influences on the global climate by introducing strong radiative effects, probably compensated by the increased greenhouse gas concentrations of past epochs.

2. ON THE EFFECTS OF THE EVOLUTION OF MICROBIAL MATS AND LAND PLANTS ON EARTH

In the second part of this thesis we extend the work presented in the first part by using spectroscopy instead of photometry to study the radiation reflected by Earth as a function of the planet’s rotation. Here, spectroscopy allow us to retrieve more information than photometry since the former allows us to identify surface signatures related to the presence of life in the spectrum of our planet. As we were interested in exploring the possible effects that the evolution of life over land might have had on the way our planet looked from afar, here we concentrated on the Earth 500 Ma ago, as this is approximately the time when advanced land-plants evolved. To study this effect on the VIS-NIR disk-integrated Earth spectra we explored several scenarios which go from bare surfaces, to continents covered by microbial mats, to continents inhabited by evolved plants.

To do that, we used a radiative transfer model to generate a 1-dimensional synthetic spectral library that covers a wide range of viewing and incident angles, surface and cloud types, and atmospheric composition, in order to be able to model a variety of planet scenarios. We also present a disk-integrated code that allows us to calculate disk-integrated spectra for any viewing geometry, surface map and cloudiness distribution, by using the aforementioned spectral database.

We found that the evolution from bare surfaces, to the colonization of microbial mats over land, to the evolution of high plants over continents, produces detectable changes in the disk-integrated reflected light of Earth 500 Ma ago, that vary substantially as Earth rotates. By convolving these simulated spectra of Earth 500 Ma ago against standard astronomical filters, we were able to conclude that using photometric observations of an Earth-like planet in different filters may allow us to discriminate between a planet whose continents are covered by deserts, vegetation or large extensions of microbial mats.

3. CHARACTERIZING THE PURPLE EARTH

In the last part of this thesis we extended the work presented in the previous part by exploring the effect of considering a different atmospheric composition. As our planet seems to have been inhabited for at least 85% of its history, here instead of using the present-day atmosphere of Earth for the simulations of the globally-averaged views of our planet, as a study case, we focus on the Earth 3.0 Ga ago. At that time, oxygen was negligible in the atmosphere and CO2 and CH4 were much more abundant. This particular period in history is also interesting due to the presence of purple bacteria as the dominant form of life over the Earth’s surface. These type of bacteria are photosynthetic microorganisms that convert sunlight into chemical energy through anoxygenic photosynthesis and are among the first forms of life on Earth, they are thought to have appeared 3.8 Ga ago.

Thus, here we studied the possibility of detecting primitive life forms, such as purple bacteria, on the disk-integrated spectra of Earth, taking into account different levels of cloud cover, continental distributions, and different scenarios where purple bacteria could be found. To conduct this study we had to take measurements of the reflectance spectrum of such bacteria in the laboratory. To do that, we used cultures of Rhodobacter non-sulfur bacteria grown as a suspension of cells in a liquid media. We found that purple bacteria have a reflectance spectrum which has a strong reflectivity increase similar to that of leafy plants, the red-edge, although shifted redwards.

We found that the presence of purple bacteria could be detected in the disk-integrated spectra depending on how spread they are over the planet. If purple bacteria are found in oceans and over continents, its spectral feature is detectable in both cloud-free and cloudy scenarios. Whereas if purple bacteria are located only over oceans they are almost undetectable in the cloudy cases. When considering a more realistic case where purple bacteria are found only in coastal zones, purple bacteria’s signature can be detected in both cloud-free and cloudy atmospheres, although the signal is smaller in the last case.

Finally, by convolving the reflectance spectra of Earth 3.0 Ga ago against standard astronomical filters and comparing this result with that of Earth 500 Ma ago, we concluded that using photometric observations in different filters may allow us to distinguish between a present-day Earth whose continents are covered by deserts, vegetation or microbial mats, from an Achaean Earth where purple bacteria have colonized the planet.

REFERENCES

Barclay et al. 2013, ApJ, 768, 101

Borucki et al. 2013, Science, 340, 587

Brogi et al. 2012, Nature, 486, 502

Cassan et al. 2012, Nature, 481, 167

Cockell et al. 2009, Astrobiology, 9, 623

Cowan et al. 2009, ApJ, 700, 915

Des Marais et al. 2002, Astrobiology, 2, 153

Gilliland et al. 2013, ApJ, 766, 40

Hegde & Kaltenegger 2013, Astrobiology, 13, 47

Kaltenegger et al. 2007, ApJ, 658, 598

Kiang et al. 2007a, Astrobiology, 7, 252

Kiang et al. 2007b, Astrobiology, 7, 222

McKay & Smith 2005, Icarus, 178, 274

Meadows, V. S. 2006, in IAU Colloq. 200 (Cambridge University Press), ed. C. Aime & F. Vakili, 25-34

Mojzsis et al. 1996, Nature, 384, 55

Montañés-Rodríguez et al. 2006, ApJ, 651, 544

Pallé et al. 2004, Science, 304, 1299

Pallé et al. 2003, J. Geophys. Res. (Atmos.), 108, 4710

Parkinson et al. 2008, Orig Life Evol Biosph, 38, 355

Robinson et al. 2011, Astrobiology, 11, 393

Sanromá & Pallé 2012, ApJ, 744, 188

Sanromá et al. 2013, ApJ, 766, 133

Sanromá et al. 2014. ApJ, 780, 52

Seager at al.2005, Astrobiology, 5, 372

Seager et al. 2012, Astrobiology, 12, 61

Sterzik et al. 2012, Nature, 483, 64

Spohn & Schubert 2003, Icarus, 161, 456

Swain et al. 2009, ApJ, 704, 1616

Tinetti et al. 2006, ApJ, 644, L129

Tinetti et al. 2007, Nature, 448, 169

Turnbull et al. 2006, ApJ, 644, 551

Abstract

The search for extrasolar planets has become one of the most exciting fields of astrophysics in the last years. Since the discovery of the first planet orbiting a main sequence star other than the Sun in 1995, the number of detected exoplanets has increased exponentially. Although most of the discovered planets are gas giants, in the last few years we have been able to detect smaller exoplanets, some of them being super-Earths that orbit near or even within the habitable zone of their stars.

Moreover, the detection of some Earth-size and Moon-size planets has been recently announced. Therefore, it seems that the detection of an Earth-twin is just a matterof time.

The study of Earth will be essential for understanding future observations of planets similar to our own, and also for the characterization and the search forlife elsewhere. In the last years, several studies both theoretical and observational have been carried out with the aim of determining how our planet would look like to an extrasolar observer. In this thesis we extend these studies to time scales of millions of years, throughout the Earth evolution, by studying how different continental distribution, cloudiness levels, atmospheric compositions and/or the evolution of different life forms, could have affected the photometric and spectroscopic characteristics of Earth.

Clouds, one of the most important parameters in the energy balance of our planet, are tied to ocean currents and orography. Thus, in the first part of this thesis we have attempted to study the Earth’s large-scale cloudiness behavior with the goal of using these knowledge to reconstruct the possible cloud distribution in past epoch of our planet. Albedo, which is intrinsically related to cloudiness, is one of the most important parameters in the energy balance of our planet since it directly controls the temperature. In this thesis we have used a simple albedo model in order to study the photometric variability of our planet along its history, when the continental distribution was very different from that which we are familiar with today.

With the objective of extending the photometric study carried out in the first part of this thesis, in the second part we have studied the spectroscopic characteristics of our planet. To do that, we used a radiative transfer model with the aim of generating a 1-dimensional spectra database which covers a wide range of surface types, incident and observing angles, clouds, aerosols, and atmospheric compositions, in order to be able to calculate disk-integrated views of our planet for a variety of geometries.

Finally, this spectral database was used to study the possible effect that the appearance of life over continents and oceans could have had on the Earth’s spectrum. Throughout the Earth’s evolution, several events such as the colonization of purple bacteria – one of the first photosynthetic life forms that colonized our planet – or the evolution of microbial mats and land plants over continents. In this thesis we show that the presence of different forms of life over continental surfaces, could be detected and characterized by studying the sunlight reflected by our planet.

UNESCO CODES: 2104.03, 2104.07

Índice

1 Introduction 1

1.1 The road to exoplanets . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.2 Exoplanet detection methods . . . . . . . . . . . . . . . . . . . . . . 2

1.2.1 Radial velocity . . . . . . . . . . . . . . . . . . . . . . . . . . 2

1.2.2 Transits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

1.2.3 Astrometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

1.2.4 Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

1.2.5 Microlensing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

1.3 Discovered exoplanets so far . . . . . . . . . . . . . . . . . . . . . . . 7

1.4 The Earth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

1.4.1 Earth’s energy budget . . . . . . . . . . . . . . . . . . . . . . 11

1.4.2 Earth’s albedo . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

1.4.3 Clouds on Earth . . . . . . . . . . . . . . . . . . . . . . . . . 14

1.5 The search for life: What makes a planet habitable? . . . . . . . . . . 16

1.5.1 Habitability within the Solar System? . . . . . . . . . . . . . . 17

1.5.2 Current potentially habitable exoplanets . . . . . . . . . . . . 19

1.5.3 Biosignatures . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

1.6 The Earth as an exoplanet . . . . . . . . . . . . . . . . . . . . . . . . 23

1.7 The Earth through time . . . . . . . . . . . . . . . . . . . . . . . . . 27

1.7.1 Plate tectonics . . . . . . . . . . . . . . . . . . . . . . . . . . 28

1.7.2 The origin of life . . . . . . . . . . . . . . . . . . . . . . . . . 29

1.7.3 The faint young Sun problem . . . . . . . . . . . . . . . . . . 31

1.7.4 The Great Oxidation Event . . . . . . . . . . . . . . . . . . . 32

1.8 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

2 Model Inputs 35

2.1 Atmospheric profiles . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

2.2 Earth’s continental distribution along time. . . . . . . . . . . . . . . . 37

2.3 Cloud distribution and properties . . . . . . . . . . . . . . . . . . . . 39

2.3.1 A semiempirical model for Earth’s clouds . . . . . . . . . . . . 39

2.3.2 Cloud reconstruction . . . . . . . . . . . . . . . . . . . . . . . 45

2.4 Optical properties of aerosols and clouds . . . . . . . . . . . . . . . . 47

2.5 Spectral albedo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

2.5.1 Purple bacteria spectral albedo . . . . . . . . . . . . . . . . . 49

3 Constructing a disk-integrated Earth spectrum model 53

3.1 A simple reflectance model . . . . . . . . . . . . . . . . . . . . . . . . 53

3.1.1 Example model outputs: Test planets . . . . . . . . . . . . . . 54

3.2 Spectral Earth model . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

3.2.1 Radiative transfer model . . . . . . . . . . . . . . . . . . . . . 58

3.2.2 Disk-integrated spectrum of Earth . . . . . . . . . . . . . . . . 64

4 Reconstructing the Cloud Distribution and Light Curves of Earth

Along its History 73

5 On the Effects of the Evolution of Microbial Mats and Land Plants

on Earth 83

6 Characterizing the Purple Earth 93

7 Results and Conclusions 107

A Annex A 111

Conclusiones

for the characterization of future detected Earth-like exoplanets. We have studied how different atmospheric composition, continental and cloud distribution, and the evolution of different life forms might have affected the way our planet looked from afar. In this Chapter we briefly summarize the main results and conclusions resulting

from this work.

Models:

We used a simple albedo model in order to study the photometric variability of Earth at different epochs according to the different cloud and geographic distribution that Earth has shown along its history. This model calculates the

visible reflected light curves of our planet for a given day, using as input data cloud and snow/ice cover maps, and surface maps. With this information the model is able to determine the Bond albedo at each time of the day and in any direction.

A radiative transfer code was used to generate a 1-dimensional synthetic spectral library with the aim of being able to simulate the Earth’s reflectance spectra for a wide range of scenarios and geometries. We produced more than 7500 synthetic spectra that cover a range of insolation and reflection angles, nine surface types, three cloud types, and seven atmospheric profiles, six of them corresponding to the composition of the present-day atmosphere and the other one corresponding to that of early Earth. These spectra cover both the VIS-NIR and the IR.

We also developed an IDL code which calculates the disk-integrated spectrum of our planet for a given viewing geometry and resolution. This code uses as input data surface and cloudiness maps, and the spectral library discussed above.

Cloud Reconstruction:

We used satellite-based estimations of the distribution of clouds taken from the ISCCP as well as information about the surface properties and continental distribution of Earth in order to study the large-scale cloudiness behavior of our planet. By using these measurements we obtained empirical relationships between cloud cover and its location over Earth’s surface depending on latitude and the underlying surface types.

However, there are some small-structures mainly over oceans that our models is not able to reproduce. These differences are due to oceanic effects, such us sea surface temperature, trade winds, ocean currents, and regional meteorological phenomena such as «El niño»/»La niña», which can influence cloud formation considerably.

The empirical relations between cloud cover, longitude, and surface type were also used in order to reconstruct the possible cloud distribution of past epochs of Earth such as 90, 230, 340, and 500 Ma ago, when plate tectonics has changed the appearance of our planet.

Daily Light Curves:

We modeled the 24-hour Bond albedo variations of present Earth over the course of a day for both modeled and measured cloud cover, using the aforementioned albedo model. We found a good agreement between the shape of the light curve obtained from reconstructed clouds and that calculated from measured cloud cover. However, the mean value of the light curve obtained from the reconstructed cloudiness tends to be higher than the one obtained from real cloud cover. These discrepancies in the Albedo variability are owing to the fact

that reconstructed cloud cover presents much less changes in the longitudinal direction than the real one, affecting the Bond albedo values.

We also studied the 24-hour albedo variations of the past epochs of Earth: 90, 230, 340, and 500 Ma ago. The first three epochs presented similar daily variations, with a smooth enhancement in reflectivity due to the presence of continental masses. These similarities are related to the fact that these epochs have a similar ocean-land-vegetation distribution, where continents are almost totally clustered and located mainly in the Northern hemisphere. Whereas, the light curve of Earth 500 Ma ago presented more variability than that of the

other epochs. This is again related to the continental distribution. In this case, continents were clustered in one hemisphere but it also had some big islands that caused strong variability in the light reflected by the planet. Moreover, this epoch showed a higher mean albedo which is due to the fact that continents

were completely desert.

We concluded that the mean albedo value of Earth has remained approximately constant along its recent history, around 0.3. However, daily variations in the past may have been larger than present day’s. In particular, daily variation of Earth 500 Ma ago were three times larger than present’s day. This increased variability could help in the determination of the rotational period of the planet seen from an astronomical distance. Moreover, these higher mean albedo values should have had profound influences on the global climate by introducing strong radiative effects, probably compensated by the increased greenhouse gas concentrations of past epochs.

Laboratory Measurements:

As one of the inputs of the radiative transfer model is the reflectance albedo of the different surface types, two different experiments were carried out in order to get the albedo spectrum of purple bacteria. These measurements were taken in the spectral range 0.3-2.5 um. We found that purple bacteria have a reflectance

spectrum which has a strong reflectivity increase similar to that of leafy plants, the red-edge, although shifted redwards.

Spectroscopy:

By using both the disk-integrated code and the spectral library, we simulated the VIS-NIR disk-averaged spectra of our planet at different historical epochs such as 90, 230, and 340 Ma ago, for a cloud-free atmosphere and for a cloudy one, and assuming that the atmospheric composition was the same as that of present-day Earth. We found that, in the case of considering the clear sky condition, the shape of the spectra is very similar for these three cases. However, the amplitude of the daily variations is different. The addition of clouds into the model results in the increase of the reflectivity of the planet and reduces dramatically the albedo variability over the course of a day. Hence, these differences between the albedo of different epochs are almost diluted, being very difficult to distinguish among them. We concluded that changing the continental distribution does not have a high impact in the globally-averaged spectral variations.

The VIS-NIR globally-integrated spectral variability of Earth 500 Ma ago was also modeled. Aiming at exploring the effect that the evolution of life over land might have had on the spectra of our planet, we considered five different scenarios regarding the continental surface properties: bare continents (deserts), vegetated continents, continents covered by microbial mats, and continents covered by a mixture of microbial mats and deserts/vegetation. We also included the contribution of clouds in our simulations.

We convolved the simulated spectra of Earth 500 Ma ago against standard astronomical filters. We concluded that using photometric observations of an Earth-like planet in different filters may allow us to discriminate between a

planet whose continents are covered by deserts, vegetation or large extensions of microbial mats.

The VIS-NIR spectra of Earth during the Archean (3.0 Ga ago) was also modeled. During this time the Sun was just ~79% as bright as it is today, CO2 and CH4 were abundant in the atmosphere, and O2 was negligible. We

explored how a different atmospheric composition and early forms of life, such as purple bacteria, could have affected the spectra of our planet. We calculated these spectra taking into account three different scenarios: one where purple bacteria have colonized the whole planet (both continents and water), another where purple bacteria are found only in oceans, and finally a third one were purple bacteria are located over coastal regions.

cloudy atmospheres, although the signal is smaller in the last case.

Finally, by convolving the reflectance spectra of Earth 3.0 Ga ago against standard astronomical filters and comparing this result with that of Earth 500 Ma ago, we concluded that using photometric observations in different filters may allow us to distinguish between a present-day Earth whose continents are covered by deserts, vegetation or microbial mats, from an Archean Earth where purple bacteria have colonized the planet.