Looking back at 2013. Part I – A fresh water lake on Mars

The American news channel CNN have selected what they consider to be the ten most important news about natural science and space in 2013. Gladly, planetary systems research top the list, with four different news. Other news concern fundamental physics (two), paleoanthropology and biology (two), cosmology (one) and climate (one). The four planetary news are about a fresh water lake on Mars, the spacecraft Voyager I that is leaving the Solar System, the space telescope Kepler looking for exoplanets, and the spectacular meteorite impact in Chelyabinsk, Russia. In four posts I will comment on each of these four top news of 2013.

This first post is about the finding of an ancient dried out fresh water lake on the planet Mars. First I will describe exactly what has been found, and in what way. This is necessary in order to understand what traces a lake leaves behind, billions of years after it has disappeared, and what it takes to actually identify and record these tracks. Then I will describe how these findings have been interpreted, and what that interpretation really means. The interpretation is an example of detective work – the ability to reconstruct a sequence of events from a few damaged, faded and indistinct pieces of a puzzle.

On November 26, 2011, NASA launched the space probe Mars Science Laboratory (MSL). Almost exactly three years earlier, when I worked at the Jet Propulsion Laboratory (JPL) in Pasadena (CA), I had the opportunity to see how the spacecraft was assembled and tested. The picture below was taken at that time – note especially the engineer at the lower left for scale. Between the white casing and black cone-shaped heat shield the rover Curiosity is located, which now roams the martian surface since the landing on August 6, 2012. Curiosity landed inside a crater named Gale, which has a diameter of 154 km, near the foothills of the central mountain Aeolis Mons. The first stop was an area called Yellowknife Bay, located only 450 meters from the landing site, carefully selected in advance through the study of the Mars surface from various orbiting spacecraft. The place in question has been shown to consist of three geological layers or strata, each about 1.5-2 meters thick, which have been exposed as a stair-like formation by erosion. With increasing depth they are informally known as Glenelg, Gillespie Lake, and Sheepbed. It is the bottom layer that is the cause of all the commotion. In a number of papers published in Science on December 9, 2012, the 400 person strong MSL team describe why they think Sheepbed represents sediments that were formed when fine rocky dust sank to the bottom of a quiet lake long ago.

Mars Science Laboratory during mounting and testing at the Jet Propulsion Laboratory in November 2008. Private photo.

Mars Science Laboratory during mounting and testing at the Jet Propulsion Laboratory in November 2008. Private photo.

Description of the finding

Size, type and appearance

Curiosity has only rolled about 60 meters across Sheepbed’s surface, but observations made by orbiting spacecraft show that the exposed portion of the layer extends over the entire four square-kilometer Yellowknife Bay, and possibly beyond that over a 30 square-kilometer area. If true, the lake has been at least six kilometers across.

Below a very thin surface layer Sheepbed has an unusual light gray color and consists of very small grains, cemented together into compact rock. The size of the grains has been investigated with a special camera called Mars Hand Lens Imager (MAHLI), with a maximum resolution of 14 micrometers per pixel, which means that it is capable of distinguishing single grains that are a few hundredths of a millimeter across. The current images do not reach that resolution, but one can still conclude that the grains in the material have sizes that are just below 50 micrometers. This is called a mudstone, distinguishing it from the finer claystone and the coarser sandstone. Drilling in the material confirms that it has a weak to moderate hardness typical of terrestrial mudstone and is softer than sandstone. The layer above, Gillespie Lake, instead consists of sandstone with coarser grains of very different sizes, often having angular shapes, which are well mixed with no hint of size sorting.

Sheepbed has been exposed due to the partial eroding of the overlaying stratum Gillespie Lake. Image Credit: NASA/JPL-Caltech/MSSS

Sheepbed has been exposed due to the partial eroding of the overlaying stratum Gillespie Lake. Image Credit: NASA/JPL-Caltech/MSSS

The Sheepbed mudstone is not entirely homogeneous, but contains structures called nodules, hollow nodules, raised ridges, and filled fractures. Nodules are spherical concretions (particularly hard rock) that reveal their existence as local elevations on the surface, as the wind mainly has eroded the surrounding softer rock. Brushing confirm that they are unusually hard. Around 4500 nodules were examined and they have sizes between 0.4 and 8.2 millimeters, with a mean of 1.2 millimeters. They cover about 2% of the surface. Hollow nodules are spherical void spaces in the mudstone, with reinforced walls. Around 1200 investigated hollow nodules have sizes between 0.6-5.6 millimeters, with a mean of 1.2 millimeters. They cover about 4-8% of the surface. Raised ridges are narrow, straight or slightly curved, with a length of several centimeters, an average width of about three millimeters, and a height exceeding the width. They often intersect each other at right angles. Nodules and hollow nodules occur mixed together, but these structures are extremely rare in areas where there are raised ridges.

Filled cracks are long and thin channels that crisscross through the mudstone, both horizontally and vertically, and are filled with a light material. They can grow up to 30 centimeters long, sometimes being as narrow as a human hair, sometimes up to a centimeter thick, while the average diameter is about two millimeters. They cover up to 14% of the surface. In some places filled cracks cut through the hollow nodules, and the latter are then also filled with the bright material.

A hole drilled into Sheepbed on Mars, revealing the bright and fine-grained mudstone and even brighter cracks filled with calcium sulfate. Image Credit: NASA/JPL-Caltech/MSSS

A hole drilled into Sheepbed on Mars, revealing the bright and fine-grained mudstone and even brighter cracks filled with calcium sulfate. Image Credit: NASA/JPL-Caltech/MSSS

Chemical composition

While photographs provide much information about the appearance and texture of a surface, a chemical and mineralogical investigation is required to reconstruct its history. Which chemical elements are present and in what relative proportions? How have these elements been combined to form molecules and minerals? Curiosity has two instruments that are suitable for elemental analysis – APXS and ChemCam. APXS carries a radioactive substance, curium-244, which decays into plutonium-240 while emitting a helium nucleus, also called an alpha particle. The instrument is placed a few centimeters from the rocky surface one wishes to examine, and an area with a size of a few square centimeters is bombarded with alpha particles. The atomic nuclei in the rock are surrounded by electrons moving on specific orbits or shells. During collisions with alpha particles, the electrons are raised to higher shells. As they spontaneously decay back to their original shells, X-rays are emitted at specific wavelengths that are unique for the element. By observing the emitted X-rays with a spectrometer one can thus determine which elements are present and in what quantity. Hence the name APXS which stands for Alpha Particle X-ray Spectrometer. Detection of abundant elements require irradiation and observation for ten minutes, while trace elements may require several hours of observation. The technique of using X-ray spectroscopy to reveal the presence and abundances of a large number of elements at a single measurement during bombardment with heavy particles was demonstrated for the first time (using protons) by the Swedish nuclear physicists Thomas B. Johansson, Roland Akselsson and Sven A. E. Johansson in 1970, then active at the Faculty of Engineering at Lund University.

ChemCam consists of an optical instrument for photography, and LIBS that shoots a laser beam onto a target that can be up to eight meters away. The laser vaporizes the stone by breaking up the minerals into their atomic constituents. The liberated atoms emit ultraviolet radiation, visible light and infrared radiation at wavelengths that are specific to each element. This radiation is measured with a spectrometer, allowing the presence and abundance of elements to be determined, within the targeted area that is half a millimeter across. The acronym LIBS stands for Laser Induced Breakdown Spectroscopy .

The abundances of the various elements in the rock are presented and compared in a somewhat special manner. One must first realize that all forms of rock and stone, on Earth as well as on Mars, are dominated by one particular chemical element – lo and behold – oxygen, denoted by the letter O from its Latin name oxygenium. Oxygen is simply the most abundant element by number, both on Earth and on Mars. By this I do not mean the molecular oxygen that we breathe in Earth’s atmosphere (molecular oxygen O2 consists of two oxygen atoms), but the entire terrestrial content of atoms – here oxygen dominates in terms of number. Other common chemical elements in rock, such as iron (Fe after the Latin word ferrum), silicon (Si after the Latin word silicium) and magnesium (Mg) bind very tightly to this oxygen in a mineral. In fact, these bonds often remain even when rock is melted – in magma, atoms of iron, silicon and magnesium do not swim around independently, but they still hang on to one or more oxygen atoms, forming units such as FeO, SiO2, and MgO. This is why it is so difficult to extract iron from ore – it is not sufficient to melt the ore, one must also cleave the FeO molecule. For these reasons it is customary not to compare the amounts of iron, silicon and magnesium directly, but rather focus on concentrations of FeO, SiO2, and MgO in the rock (expressed as a percentage of the total mass – or weight percent, with the symbol wt%).

APXS and ChemCam show that Sheepbed, in descending order, is richest in SiO2, FeO, MgO, Al2O3, CaO and SO3, with abundances of 5 to 40 wt% each. The most common elements apart from oxygen is thus silicon, iron, magnesium, aluminum (Al), calcium (Ca), and sulfur (S). Furthermore, Na2O, Cl, TiO2, and P2O5 have concentrations above 1 wt%, which means that also sodium (Na), chlorine (Cl), titanium (Ti) and phosphorus (P) are fairly common. Finally, there is K2O, Cr, MnO, Zn, Ni and Br in small concentrations (a few hundredths or tenths wt%), indicating that potassium (K), chromium (Cr), manganese (Mn), zinc (Zn), nickel (Ni) and bromine (Br) are trace elements in the mudstone.

This rich set of detailed data can now be compared to other places on Mars. It then turns out that the composition is fairly close to the average on Mars – Sheepbed therefore does not differ significantly from the Mars surface at large. However, the amount of Al2O3 compared to TiO2 tends to be somewhat low, and also sulfur is on the low side, while the amount of chlorine is slightly elevated. Gillespie Lake is also close to average, but the top layer Glenelg differs from the others by having smaller amounts of iron and magnesium (FeO, MgO) and higher levels of potassium (K2O). There are also differences in trace elements – Glenelg has elevated levels of zinc and chromium, as well as a lower content of nickel compared to Sheepbed. Nodules do not differ dramatically from other materials in Sheepbed – at most 10% for the common elements. However, raised ridges exhibit elevated levels of iron, magnesium, chlorine and lithium (Li). The bright material that fills out cracks has elevated levels of calcium, sulfur, hydrogen (H) and strontium (Sr) . All of these similarities and differences are important clues which can be used to unravel the story of Sheepbed’s origin and evolution.

Mineralogical composition

Such a historiography requires more concrete information about the mineralogical composition of Sheepbed. APXS and ChemCam show what building blocks are available, but say nothing about how these building blocks are joined. It is like looking at neat piles of brick, wood, steel, and glass – but how did the house actually look like before it was demolished? Which minerals and rocks are the Sheepbed grains made of?

This question is much more difficult to answer than to measure the chemical composition, especially with the limited resources of a spacecraft, which cannot be compared to well-equipped laboratories on Earth. However, there is an instrument on Curiosity that come a long way, namely CheMin. The instrument, whose name is an abbreviation of Chemistry and Mineralogy employs a technique called X-ray diffraction. It relies on the fact that many minerals are crystals, meaning that its atoms are placed in definite and very precise geometric patterns. CheMin has an X-ray tube in which electrons are accelerated to high speed and collide with a plate made of cobalt. The cobalt then emits X-rays, which illuminates a rock sample in the form of dust from drilling, which has been brought into Curiosity’s interior. On its way through the sample, the X-rays are bent, much like light passing through a lens. This is done in a manner that depends strongly on the crystal structure of the minerals present in the sample. The X-rays will be divided into a number of concentric rings with different intensity and distance from the center – a so-called diffraction pattern that can be photographed. The practical difficulty lies in recognizing the pattern that specific minerals give rise to, a task that can be very difficult when a large amount of minerals coexist in the same sample, of which some may have an unusual structure and odd crystal pattern. It is also problematic that some substances (called mineraloids) are amorphous, so that the small basic molecular units are completely unordered and not arranged in nice crystal patterns (glass is an example of an amorphous material), and they cannot be easily identified since they do not give rise to unique diffraction patterns.

These limitations notwithstanding, we have a fairly good idea of Sheepbed’s mineralogy. Around 30% are unidentifiable amorphous materials, possibly containing allophane, a mineraloid consisting of Al2O3, SiO2 and water molecules (H2O). About 22% is composed of plagioclase, a mineral mixture that, together with orthoclase, forms one of the most common type of rock found on the surface of Earth, feldspar. Plagioclase consists of a mixture of two different minerals in some proportion – sodium-rich albite (NaAlSi3O8) and calcium-rich anorthite (CaAl2Si2O8). In Sheepbed, the albite and anorthite are roughly equally common. Around 20% are phyllosilicates. The name refers to the crystal structure which consists of thin parallel layers – the Greek word phyllon means “leaf” (the same word is used in, e.g., chlorophyll). The layers of so-called dioctahedral phyllosilicate consist of aluminum, silicon and oxygen, where atoms are linked in a flat mesh in a specific geometric pattern. I trioctahedral phyllosilicate, magnesium, silicon and oxygen are forming the layers instead – the phyllosilicate in Sheepbed is mainly of the latter type. These layers have the ability to bind large amounts of water between them. The dominant phyllosilicate in Sheepbed seems to be saponite which has the general formula Ca0.25(Mg,Fe)3((Si,Al)4O10)(OH)2·n(H2O).

The Earth’s mantle is composed mostly of the silicates pyroxene and olivine, which are also commonly found in meteorites. Pyroxene is a rock made up of three minerals in different proportions – enstatite (MgSiO3), ferrosilite (FeSiO3) and wollastonite (CaSiO3). In Sheepbed, about 10% of the matter is pyroxene, mainly in the form of a mixture with the composition Mg1.1Fe0.8Ca0.1Si2O6 called pigeonite, and another with the composition Ca0.8Mg0.7Fe0.5Si2O8 called augite. Olivine is a rock composed of two minerals in various proportions – forsterite (Mg2SiO4) and fayalite (Fe2SiO4). Low-iron olivine make up some 3% of the rock in Sheepbed. There is also 4% magnetite (Fe3O4) and a couple of percent calcium sulfate in the form of anhydrite (CaSO4), bassanite (2CaSO4·H2O) and gypsum (CaSO4·2H2O). Calcium sulfate is particularly concentrated in the filled cracks and is responsible for their bright color. Presence of bassanite and gypsum (such calcium sulfate is called hydrated because of its content of water) has also been observed by the camera MastCam, that has an appropriate set of filters to detect hydrated minerals. Note again the abundance of oxygen atoms in all these minerals – this is typical for all forms of rock, whether it is your backyard or on the surface of Mars.

This blend of abundant silica (SiO2) and iron, along with significant amounts of plagioclase, pigeonite and augite is a familiar mixture for geologists. It shows that the material in Sheepbed originally consisted of relatively iron-rich basalt, an igneous rock of volcanic origin. Thus, we are dealing with a volcanic ash or volcanic rock that has eroded into dust, long before it ended up on this site. These original minerals are called primary ones. An unusually low olivine abundance compared to regular basalt, and the vast array of minerals that rarely or never are found in fresh basalt – phyllosilicates, allophane and magnetite (and very small amounts of akaganeite FeO(OH)) – shows that the material has changed with time. Such minerals formed over time by various processes are called secondary minerals. These particular secondary minerals are formed by a very special process – the basalt has been exposed to liquid water.

Water and organic substances

The mudstone thus contains water – not in the sense that it is wet, but the water molecules are baked into the very crystal structure of many minerals. The question is how much water there is in the material. Curiosity also has an instrument for investigating such things – Sample Analysis at Mars, abbreviated SAM. Here, powder from drilling is placed in an oven where the sample is gradually heated to high temperatures. Phyllosilicates and similar water-rich minerals will then give off water and other volatiles, which can be identified through a variety of techniques. For example, mass spectrometers, which measure the mass of the gas molecules, can identify the type of gas molecules and measure their concentrations. The temperature at which outgassing occurs, provides important information about the type of mineral that is the source of the emitted gas in question.

In these experiments it was found that samples from Sheepbed, upon heating, mainly give off water (H2O), but also molecular hydrogen (H2), carbon dioxide (CO2), sulfur dioxide (SO2), nitrogen oxide (NO) and molecular oxygen (O2). The amount of water in the mudstone corresponds to 2.5 weight percent, and 70% of the vapor is emitted at 160°C while the remainder is outgassed at 725°C. At low temperature, the water comes from saponite, bassanite and akaganeite. At high temperature, it is particularly strongly bound water in saponite that finally is released.

Oxygen release starts at 150°C, and constitutes up to 1wt% of the mudstone. The amount of oxygen in different samples varies, in a manner that coincides with the variation of chlorine content in the material. Furthermore, oxygen release is accompanied by another gas, hydrogen chloride (HCl) which, if dissolved in water, is called hydrochloric acid. One can therefore conclude that the source of oxygen is some sort of perchlorate which all have in common that they contain the perchloride ion ClO4, for example iron percholrate (Fe(ClO4)2). This also gives an indication of the way in which chlorine is incorporated in the mudstone grains.

One measurement has attracted much attention because it is directly related to the eternal questions – did Mars once have habitable environments, and did it ever host living organisms. A sudden and temporary decline in the production rate of oxygen during heating, which happens to correspond to a sudden production of carbon dioxide, is possibly caused by the combustion of organic material. The combustion releases carbon (C), which reacts with molecular oxygen (O2) to form carbon dioxide (CO2). Unfortunately, it is very difficult to determine whether this organic material comes from the mudstone itself, or if there is a contamination from the spacecraft. The levels are very low. After extensive tests, it is believed that the spacecraft provides only a few percent of the measured carbon, although it is difficult to rule out a larger contribution. If carbon comes from the mudstone itself it remains to find its source. Possibly the carbon partly originates from meteoric material in the mudstone – it is well known that meteorites contain organic substances and they constantly rain down on the surface of Mars. The source can also be small amounts of carbonate, such as calcium carbonate (CaCO3), which is the main ingredient in limestone. All such abiotic carbon sources must first be ruled out before one can begin to speculate about more exotic options.


In laboratories on Earth, one can measure the age of terrestrial rocks, meteorites, and samples from the Moon which were brought here by Soviet unmanned space probes and American Apollo astronauts. This is done by measuring the abundances of various isotopes corresponding to different types of well-known radioactive decay sequences. Age estimates of other planet or satellite surfaces are typically done by counting craters – the more craters per square kilometer surface, the higher the age. Radiometric dating of lunar rocks is used to translate a crater count into an actual age measured in billions of years. Based on such a crater counts it is believed that Gale Crater was formed about 3.5-3.7 billion years ago. Individual fields within the crater have younger ages, which shows that some of them formed 2.9-3.5 billion years ago. As a comparison, the Solar System age is 4.57 billion years.

However, for the first time ever, Curiosity carries equipment that makes it possible to measure the age of martian rocks in situ, i.e., in place. APSX is used to measure the abundance of potassium (K), which primarily consists of the stable isotopes potassium-39 and potassium-41. A small fraction, however, consists of the radioactive isotope potassium-40, which decays into argon-40 with a half life of 1.3 billion years. Argon is a noble gas that does not occur naturally in any form of mineral, and therefore was never added to Mars from the environment in which the planet formed. All argon found on Mars today, encased deep within the mineral crystal structure, has been formed there by the radioactive decay of potassium-40. This argon is outgassed in SAM when samples are heated and can be measured. The measured concentrations of potassium and argon in a rock sample can then be used to calculate the age of the sample.

One finds that the age of the material in Sheepbed is at least 3.9 billion years, with 68% confidence. One could also say that it must be at least 3.6 billion years old, with 95% confidence. To understand what this age really means, we must track the material over time and point out a few important events in its history.

At the time when the material constituting Sheepbed today was molten lava, it contained no argon-40 because the gas easily escaped. But when the lava solidified, the escape routes of the argon closed, and the element began to accumulate as potassium decayed. However, it is highly likely that the impact that formed Gale Crater removed large parts of the argon that had formed in the meantime. One can therefore say that the clock probably was reset at the time of the big impact. We also know that some of the minerals that are found in Sheepbed have been destroyed by water, that converted primary minerals to secondary minerals. That process will reset the clock again, but only for secondary minerals. The crucial question is – how much potassium is located in the primary minerals, and how much is found in the secondary ones? This partition is crucially important when we want to interpret what an age of at least 3.6 billion years actually means.

Unfortunately, Curiosity does not carry the equipment required to investigate exactly in what minerals the potassium is located. Based on past experience of potassium partition between primary and secondary minerals in meteorites from Mars, previously studied in the laboratory, and educated guess is that about 80% of the potassium is in the primary minerals. If true, one can make the following interpretation of the measurements. Primary minerals have an age of 4.21 billion years, with an error margin of 0.35 billion years. This corresponds to the time when the lava solidified, or possibly the formation time of Gale Crater. Note that the youngest possible age, 3.86 billion years, responds quite well to estimates from crater counting. If this is true, it can be concluded that the process that destroyed primary minerals and formed secondary minerals such as saponite, must have ceased at least 1.6 billion years ago. The amount of potassium decay in secondary minerals over 1.6 billion years, combined with the amount of decay in primary mineral in 4.21 billion years, will give the mudstone as a bulk, an average age of 3.6 billion years. Thus, there has been no water for 1.6 billion years, and probably much longer.

In addition to this potassium-argon dating, Curiosity has managed to make yet another age estimate. It tells us when Sheepbed saw the light of day anew, after erosion painstakingly had removed all the rock that once covered the mudstone. The method is based on the existence of cosmic rays. Cosmic rays are mainly protons and alpha particles, that are accelerated almost to the speed of light by exploding supernovae far out in our galaxy, the Milky Way, long ago. Since Mars, in contrast to Earth, lacks a protective magnetic field and only has a very thin atmosphere, cosmic radiation has free access to the planet surface. When particles hit atoms in the ground with tremendous speed, new and very rare isotopes are formed. For example, neutrons that have been knocked off in violent collisions may collide with chlorine atoms, allowing argon-36 to form. Similarly, collisions with oxygen, silicon and magnesium gives rise to helium-3, while collisions with magnesium, silicon and aluminum may also give rise to neon-21. The cosmic radiation only penetrates about a meter into the ground, which means that production of argon-36, helium-3, and neon-21 in Sheepbed only began once it had been cleared by erosion. Based on the levels of the three noble gases measured by APXS the following ages are obtained – 72 million years from argon-36 (with an uncertainty of 15 million years), 84 million years from helium-3 (with an uncertainty of 28 million years) and 79 million years from neon-21 (with an uncertainty of 24 million years). It therefore seems that Sheepbed only has been visible on the surface of Mars during the past 80 million years, which is a short time compared to its age (1000 million years is equivalent to one billion years). One may therefore wonder how much of ancient layers, with the same importance as Sheepbed, that are hidden beneath the surface of Mars today, at large or shallow depths.

The interpretation of the find

None of the measurements described above could have been done from Earth with telescopes. They cannot be made from a spacecraft that is orbiting Mars. It is not even possible to make them from a lander without mobility, unless it manages to land in exactly the right place by chance. What you need is a radio-controlled rover that can get to the places in the terrain where there is something particularly valuable to study – which you usually do not know until you actually are in place. But these vehicles are very expensive and they are extremely difficult to build – which is why there has only been four of them in human history.

The interpretation of the findings are based on experience about Earth’s geology from field studies, knowledge of mineralogy from studies in laboratories, and measurements of the chemical properties of the elements from countless experiments, painstakingly collected during centuries. It is now that each discovery, insight, and experience become valuable, though it was ever so small and seemingly unimportant when it was presented. It is therefore all knowledge is indispensable – you never know when, and in what manner, it will become useful. In 1840, when the Swedish geologist and Uppsala university chemistry professor Lars Fredrik Svanberg (1805-1878) for the first time described saponite and reported that miners had tried to use the soft spreadable material as butter – could he have imagined that this mineral would play a significant role in the exploration of the planet Mars?


But back to Sheepbed. The material itself is volcanic basalt, with a typical composition, albeit slightly iron-rich. Possibly, it is volcanic ash, but if so, it has not rained down and formed a layer during a single volcanic eruption – there are no volcanoes within a few tens of kilometers, which would be required to form a 1.5 meter thick layer. Rather, it is basalt from the rim of Gale Crater or its surroundings, that slowly has been transformed into a very fine powder by erosion, from which the smallest particles have been picked up by the wind and carried away. The fact that the material has higher iron content, lower sulfur content, and larger particle size than ordinary dust in the Martian atmosphere, suggests a nearby source with regional characteristics. This fine dust rained down over a stagnant body of water, which existed in this location almost four billion years ago. Once in the water, the particles slowly sunk to the bottom and formed a thickening layer of sediment. Mudstones extending over such a wide region, that are this thick, that consist entirely of such small particles, that completely lacks the scars and features that characterize rapidly flowing water, all point in the same direction – towards a peaceful lake.

Based on the thickness of the layer and the rate by which sediment usually build up it can be estimated that the lake existed for over a thousand years. More extreme sedimentation rates imply life times as short as a hundred years or as long as ten thousand years. The water in the lake has had a highly destructive effect on certain minerals. For example, the sediment has originally been considerably richer in olivine, but most olivine has deteriorated and been replaced by secondary minerals. The list of these secondary minerals may seem bland, not to say boring or unimportant – unless one is aware of what these minerals are saying about the properties of the water in the lake. They show that the water was fresh and had a neutral pH.

Acidity and salinity

The water in the lake is therefore dramatically different from the extremely salty and acidic water that has left traces at other locations on Mars. This is water that a human being could drink. This is water with such a high quality, that currently existing terrestrial microorganisms could live in it. To find traces of such an environment on another planet in the Solar System has never been done before – the find is simply revolutionary.

Sheepbed contains the secondary mineral magnetite (Fe3O4), which has a grayish color. In other locations on Mars the dominant iron oxide is instead hematite (Fe2O3). This is the substance that gives Mars its characteristic red color. Very small quantities of hematite is found also in Sheepbed, but the magnetite dominates completely. The questions one must ask are the following. How did the water manage to extract iron from the sediments, and when the liberated iron reacted with oxygen, why was the end result magnetite instead of hematite? Laboratory experiments show that this happens when water has low salinity and is neutral or slightly alkaline.

The low content of sulfur, the very modest excess of chlorine, and the small abundances of calcium and magnesium sulfates in the mudstone that surrounds the filled cracks also shows that the salinity of the water must have been low. Another place on Mars that demonstrably had liquid water long ago (Burns formation at Meridiani Planum, investigated by NASA’s rover Opportunity) was so rich in sulfates that the water would have been lethal to all known terrestrial organisms. Sheepbed is therefore very different from the inhospitable Burns formation.

The Burns formation also contained a mineral called jarosite (KFe3(OH)6(SO4)2) which is formed when water has a pH of 1-3, which is extremely acid. The absence of such minerals in Sheepbed, and the high content of phyllosilicates, which form most easily when the pH is near 7 (i.e., is neutral), shows that the water of the lake has not been acid.

Formation of nodules and hollow nodules

Substances leached out from the grains into the water will eventually precipitate and turn solid anew, much like dissolved salt or sugar when the water dries off. Such newly formed material builds bridges between the original grains and bind them together. Gradually, the loose pile of sediment is transformed into a cohesive cemented mass with growing material strength. The process is called lithification and took place also in Sheepbed. In some places, this lithification was more effective than in others – this has created the hard nodules. A tendency to find larger numbers of nodules in areas that are richer in akaganeite suggests that this secondary minerals plays a role in nodule formation.

The formation of hollow nodules seems to be closely related to the transformation of olivine into saponite. This is because this chemical process releases molecular hydrogen, a gas. The production of phyllosilicates, during the existence of the lake or shortly afterwards, has thus led to the formation of gas bubbles, which are still visible in the mudstone as holes.

The lake dries out and raised ridges form

At some stage, however, the lake dried up and the sedimentary mudstone was laid bare. When the overlying water column disappeared, the hydrated minerals saponite, bassanite, gypsum and akaganeite lost parts of their water as well. This greatly reduces the volume of the mudstone, which gives rise to a well known phenomenon – the surface of the drying sediment is broken up into a complex pattern of cracks.

With time, these cracks filled up with new material. This could have happened when the dry lake occasionally was flooded during shorter periods of time. The new material was richer in iron, magnesium, silicon, chlorine, bromine and lithium than the mudstone, suggesting that the water now was more salty. The material that filled the cracks gradually solidified into a substance that is harder than the surrounding mudstone. The concentration of akaganeite appears to be particularly high in the raised ridges. When the surrounding mudstone eroded away, the harder material in the cracks remained, and are now seen as raised ridges.

Typically, mudstone form cracks with a pattern that is different from that seen in Sheepbed. The angle at which cracks cross each other is smaller than the straight angles seen in Sheepbed. However, there are a few examples of terrestrial mudstones with patterns similar to Sheepbed, which seems to form when a mixture of fresher and more salty water puddles dry in parallel. Perhaps similar conditions have prevailed at Sheepbed.

The mudstone is covered by sediments

At some stage, Sheepbed has been covered by other types of sediment. Most likely, this happened during flooding events that involved rapidly flowing water, capable of bringing particles and grains with a broader spectrum of sizes. The result is the coarse-grained sediments in the sandstone of Gillespie Lake. The coarser grains and absence of sorting both points towards a rather strong flow of water. A separate event of flooding then brought the material that formed Glenelg. This material came from another source than the material in Sheepbed and Gillespie Lake – this basalt is richer in alkaline minerals (the elevated concentrations of potassium may be explained by a larger abundance of orthoclastic feldspar). As the weight on top of Sheepbed built up, it was subjected to ever increasing loads and stresses, that led to the formation of a complex network of cracks.

These cracks eventually were flushed with water, but this water had a completely different character than the water that once filled the lake. This water was salty and acidic. As a result, it left behind a mixture of minerals rich in calcium sulfate. This is the brighter material, which now fill the cracks. In places where the cracks cut through hollow nodules, the cavities have also been filled by the new material.

We do not know the thickness of the layer of rock that once covered Sheepbed. The only thing one can say is that Sheepbed did not get buried deep enough for the temperature, which increases with depth in the crust, to exceed 60-80°C. At such temperatures, saponite is converted into another smectitic phyllosilicate known as illite, which is not observed.

Therefore, the Sheepbed region has been under water on numerous occasions, but the floods have behaved differently and the quality of the water has changed from time to time.

Sheepbed lay hidden in the soil for several billion years. Large portions are probably still hidden away, under the Gillespie Lake and Glenelg strata. But gradually, the layer got exposed by the utterly slow erosion. And here we are, 80 million years later, examining this peculiar and strange, yet familiar landscape, that once was.

The importance of this finding is huge, because it says something significant about Mars. We already knew that surface water has been common on Mars early in its history. The evidence of flowing water left behind are visible in the form of dry riverbeds and deltas, but also in the form of sulfate-rich sediments that previous rovers have uncovered and explored. However, this water was very inhospitable – extremely acid and very salty. This is the first time we have found compelling evidence that fresh and neutral water existed on Mars for extended periods as well. This environment has also included a set of chemical elements, such as carbon, hydrogen, oxygen, nitrogen, potassium and sulfur that microorganisms need to survive. It is the first time we have found an environment on another planet, in which currently known terrestrial microorganisms would have been able to survive. Two critical questions that remain to be answered are the following – how common were such environments on Mars, and did they exist long enough for life to emerge on Mars as well?


Farley, K. A. et al. (2013). In situ Radiometric and Exposure Age Dating of the Martian Surface. Science Express (DOI:10.1126/science.1247166 ).

Grotzinger, J. P. et al. (2013). A Habitable Fluvio-Lacustrine Environment at Yellowknife Bay, Gale Crater, Mars. Science Express (DOI: 10.1126/science.1242777).

Johansson, T. B., Akselsson, R., Johansson, S. A. E. (1970). X-ray analysis: Elemental trace analysis at the 10-12 g level. Nuclear Instruments and Methods 84, 141-143.

McLennan, S. M. et al. (2013). Elemental Geochemistry of Sedimantary Rocks at Yellowknife Bay, Gale Crater, Mars. Science Express (DOI:10.1126/science.1244734).

Ming, D. W. et al. (2013). Volatile and Organic Compositions of Sedimentary Rocks in Yellowknife Bay, Gale Crater, Mars. Science Express (DOI: 10.1126/science.1245267).

Svanberg, L. F. (1840). Saponit och rosit, tvenne nya mineralier. Kongl. Vetenskapsacademiens Handlingar för år 1840. P. A. Norstedt & Söner, Stockholm.

Vaniman, D. T. et al. (2013). Mineralogy of a Mudstone at Yellowknife Bay, Gale Crater, Mars. Science Express (DOI: 10.1126/science.1243480).