GeoLog

Mars

Mars Rocks – introducing a citizen science project

Mars Rocks – introducing a citizen science project

GeoLog followers will remember our previous report on Citizen Geoscience: the exciting possibilities it presents for the acquisition of data, whilst cautioning against the exploitation of volunteered labour. This blog presents a Citizen Science platform that goes beyond data collection to analysis, specifically for geological changes in remote sensing imagery of Mars. Jessica Wardlaw, a Postdoctoral Research Associate in Web GIS, at the Nottingham Geospatial Institute, introduces ‘iMars’ and explains 1) its scientific mission and 2) why imagery analysis is especially suitable for a crowd sourcing approach, so that you might consider where and how to apply it to your project.

Imagine, just for a moment, that the Mars Geological Survey invited you to an interview for the position of Scientist in Charge. Why and how would you reconstruct the geological past for a remote planet such as Mars? Where would you start? Earth is the “Goldilocks” planet, not only for human habitation but for geologists too, who can sample and test rock to understand the evolution of the Earth’s surface on which to base well-established theories such as plate tectonics. To understand the geological past and processes of remote planets, however, requires different approaches.

Planetary scientists investigate the climate and atmosphere, and the geological terrain, of planets to further understanding of our own place in the solar system. Mars provides a scintillating snapshot of early Earth; whilst some scientists contend that plate tectonics has historically happened on Mars, 70% of its surface dates from the moment it formed and provides a platform from which to view Earth in its infancy. In fact, despite our limited knowledge of Mars, it has already informed our understanding of Earth, inspiring James Lovelock’s Gaia theory. Imminent missions to the red planet are also already exploiting geological information to inform landing sites and routes of roving vehicles on Mars. The more information scientists have, the more likely missions are to land in suitable locations to successfully pursue scientific goals, such as understanding the ability of the Martian environment to support life and water, both now and in the past, which could further theories on the origins of the solar system, life on Earth, and Earth’s destiny.

Many will remember this summer for the astonishing images that arrived from Pluto, but 50 years ago, almost to the day, people celebrated the first successful fly-by mission to Mars. Mariner 4 took 21 images from a distance 6,000 miles, which, after the initial excitement, disappointingly revealed that Mars had a Moon-like cratered surface, and led to a long-held misconception of a dead, red planet. It was in 1976 that two Viking landers touched down on the red soil for the first time, paving the way for further Martian missions, with the first mission of the European Space Agency’s ExoMars programme launching next year.

The first Mars photograph and our first close-up of another planet. A representation of digital data radioed by the Mariner 4 spacecraft on 15th July 1965. (Credit: NASA/JPL-Caltech/Dan Goods)

The first Mars photograph and our first close-up of another planet. A representation of digital data radioed by the Mariner 4 spacecraft on 15th July 1965. (Credit: NASA/JPL-Caltech/Dan Goods)

Scientists analyse the size and density of craters from meteorite impacts to age the surface. The theory goes that smaller meteorites collide with a planet much more frequently than larger ones, and older surfaces have more craters because they have been exposed for longer. Advances in imaging technology since then now provide scientists with greater granularity than ever before and glimpses of other geologic features, recognisable from the surface of the Earth; sand dunes, dust devils, debris avalanches, gullies, canyons all appear and tell us about the planet’s climatic processes. The Planet Four website is just one example.

The images taken of Mars over the last forty years reveal changes on the surface that indicate invaluable information that help us to understand the climate and geology of the planet. Changes are visible in imagery over a variety of timescales, from rapidly-moving dust devils (much bigger that the one that once trapped me in Death Valley), seasonal fluctuations of the polar ice caps and recurring slope lineae (recently reported to indicate contemporary water activity) polar ice caps and the snail-slow shaping of sand dunes.

Three images of the same location taken at different times over one Martian year show how the seasonal fluctuation of the polar cap of condensed carbon dioxide (dry ice), between its solid and gaseous state, destabilises a Martian dune at high altitude to cause sand avalanches and ripple changes. (Credit: NASA/JPL/University of Arizona)

Three images of the same location taken at different times over one Martian year show how the seasonal fluctuation of the polar cap of condensed carbon dioxide (dry ice), between its solid and gaseous state, destabilises a Martian dune at high altitude to cause sand avalanches and ripple changes. (Credit: NASA/JPL/University of Arizona)

The quality and coverage of these images, however, varies greatly due to atmospheric conditions and tilt of the camera amongst other reasons. To create a consistent album of imagery, that we can confidently compare and use to identify geological changes in the images, requires considerable computational work. Images from across as much of the Martian surface as possible must be processed to remove those of poor quality and correct for different coordinate systems (co-registration) and terrain (ortho-rectification).

The iMars project is applying the latest Big Data mining techniques to over 400,000 images, so that they can be used to compute and classify changes in geological features. On a Citizen Science platform, Mars in Motion, volunteers will define the nature and scale of changes in surface features from ortho-rectified and co-registered images to a much greater detail. Human performance is inherently variable in ways we cannot fully control, either, in the same way that we can control the performance of an algorithm. Although we are investigating this too, this would require another blog post! For now I will describe the reasons why we are using a crowd-sourcing approach for this project so that you might consider how you could apply it to your research.

First of all, humans have evolved over millions of years to identify subtle variations in visual patterns to a more sophisticated level than computers currently can. Computers can execute repetitive tasks and store an infinite amount of information with far less impact on their performance than humans; the human mind, however, has proved to be too flexible and creative for computers to fully replicate, with the success of Citizen Science projects such as Galaxy Zoo, which has so far resulted in 48 academic publications. The slow seasonal shift of sand dunes on Mars, for example, would require a computer algorithm of inordinate intelligence to identify, as previous attempts to automatically detect impact craters, valley networks and sand dunes in images of Mars have found. Recent research has resulted in some very sophisticated algorithms for image analysis, but detection of changes in such a range of geological features over the range of spatial and temporal scales that we are looking to do is computationally complex and expensive. Without sending somebody to Mars, how do we know whether the computer is correct? Machine learning algorithms can only calculate what you ask them to, so are ill-equipped to make the sort of serendipitous discoveries of the unknown required in the detection of change. Volunteers in the Mars in Motion project will seek differences (Figure 4), rather than similarities, between the images and it is inherently challenging to program a computer to find something that you don’t even know to look for.

Mars in Motion: Spot the difference...on the surface of Mars!

Mars in Motion: Spot the difference…on the surface of Mars!

Secondly, we have so much data that scientists could not possibly do all of this themselves! In many areas of science and humanities, but especially in Earth and Planetary observation, Big Data capture is growing at an astronomical rate, far faster than resources and techniques for its analysis can keep up so that we are increasingly unable to handle it. This is where geoscientists have started to join the trend for recruiting volunteers to analyse imagery with some success; through large crowd-sourcing image analysis projects, like TomNod, citizens continually contribute interpretation of images for social and scientific purposes. The number of volunteers, however, is finite and the increase in data places more and more demand upon their time. Researchers using the Citizen Science approach must now carefully consider how their projects can utilise volunteers’ time effectively, efficiently and ethically.

Third and finally, a crowd-sourcing approach exposes the public to improvements in imaging technology and brings the dynamic nature of the Martian surface to life. This can only improve the chances of space exploration receiving further funding and entering classrooms through the way it combines many areas of Science, Technology, Engineering and Mathematics. Serendipitously, the engagement of the public also increases the number of pairs of eyes that analyse the images and, as such, the confidence with which scientists can use their classifications. As we collect more and more data, image analysis will necessarily require collaboration between humans and computers, as well as between volunteers and researchers, to manage it.

I hope this post gives you an insight into how we are applying the Citizen Science to consider how it might help your research too. There is actually no better time to try setting up a Citizen Science project with the launch of the Zooniverse project builder, which makes it easier than ever before to build your own project.

By Jessica Wardlaw, researcher at the University of Nottingham

The research leading to these results has received funding from the European Union Seventh Framework Programme (FP7/2007-2013) under the iMars grant agreement no. 607379.

Visit www.i-mars.eu and follow @JessWardlaw for updates on iMars and Mars in Motion.

Geosciences Column: The quest for life on Mars

Geosciences Column: The quest for life on Mars

Understanding where we come from and whether Earth is the only habitable planet in the Solar System has been a long standing conundrum in science. Partly because it is our nearest neighbour, partly because of its past and current similarities with our own home, Mars, the red planet, is a likely contender in the quest for extra-terrestrial life. In this guest blog post, James Lewis, a PhD student at Imperial College London, takes a brief look at the findings of his recent research. Strap up, we are rocketing over to Mars!

Mars has always been at the forefront of our imaginations when we picture alien life and the discoveries planetary science has made in recent decades reveal that the idea of our neighbouring world having once been inhabited is not so far-fetched. Mars appears to have once been a habitable world, the question is did life ever exist there? This is one of the questions that Curiosity Rover is attempting to shed more light on but results so far have been inconclusive. One potential problem is that the mineralogy of Mars might seriously disrupt experiments looking for evidence of ancient microscopic Martians. Chlorine salts have already been proven to be problematic and in research, published today, and summarised in the following article I have shown that a salt containing iron, sulfur and oxygen, known as jarosite, can also be added to the list of problematic minerals for life detection experiments.

Eberswalde Delta on Mars, evidence for an ancient persistent flow of water over an extended period of time on the Martian surface. Image Credit: NASA/JPL/MSSS.

Eberswalde Delta on Mars, evidence for an ancient persistent flow of water over an extended period of time on the Martian surface. Image Credit: NASA/JPL/MSSS.

The satellites, landers and rovers sent to Mars have started to unravel many of the mysteries of the red planet. Perhaps their most exciting discovery is that ancient Mars may have been a habitable environment for life. The Martian surface at present is extremely cold, exceptionally dry and bombarded by ultraviolet radiation. The atmosphere is at such a low pressure that liquid water would instantly vaporise. However, characteristic landforms and the presence of minerals that we know only form in water have revealed that ancient Mars had persistent surface or near surface liquid water. The presence of liquid water is exciting because it is a precursor for life and for it to persist on the surface would require a warmer thicker atmosphere.

This potentially habitable liquid water existed billions of years ago, so how can we investigate if life ever existed in these environments? If ancient Martians existed they would likely be microscopic organisms like bacteria on Earth. We could look for the fossils they might leave behind but these features would be extremely small and there are many non-biological processes that can form similar structures. The least ambiguous evidence would be to find chemical compounds that only life leaves behind. As biological molecules contain carbon they fall under a chemical class called organic compounds. However, not all organic compounds are biological. For example, asteroids and comets contain non biological organic compounds that formed in the early Solar System.

Comets and asteroids have been impacting Mars throughout its history so when we send missions to Mars we would expect to see the organic molecules delivered by impacts from outer space. The strange thing is that we haven’t. If we can’t detect compounds we know should be there, what are our chances of detecting possible organic compounds indicative of life? All that has been detected so far are very simple organic compounds with chlorine attached. Their origin is uncertain as similar compounds are used as cleaning agents on Earth and sometimes as reagents inside the rovers, so they could just be contamination. However, recent discoveries have complicated things even further; in 2008 a mineral called perchlorate was discovered on Mars. Perchlorate is very rare on Earth as it is only stable in very arid environments such as the Atacama Desert and the Dry Valleys of Antarctica. Perchlorate has now been discovered by multiple Mars’ missions so it would appear it is widespread in the extremely arid present day Martian surface.

The Phoenix Lander made the first detection of perchlorate on Mars in 2008. Dusty Martian soil can be seen in the background and on the Lander’s frame. Image Credit: NASA/JPL-Caltech/University of Arizona/Texas A&M University.

The Phoenix Lander made the first detection of perchlorate on Mars in 2008. Dusty Martian soil can be seen in the background and on the Lander’s frame. Image Credit: NASA/JPL-Caltech/University of Arizona/Texas A&M University.

Perchlorate is a big complication in our search for organic compounds on Mars. The most common technique used to analyse samples for the presence of organic compounds is to heat materials in an inert atmosphere until organic compounds break down and go into the gas phase. The chemical composition of this gas can then be analysed. For example, on the Curiosity rover the gas passes from the sample oven into a and then a mass spectrometer, which separates out the constituent gases and identifies them. The problem with perchlorate is that it breaks down at low temperatures, in fact just at the temperatures that organic molecules would start to break down and be detectable. Perchlorate releases oxygen and chlorine when it thermally decomposes. Oxygen will react with, and break down, organic compounds into carbon dioxide and water. So it will greatly reduce the instrument’s ability to detect organic molecules if it is present in the sample heating oven. The simultaneous release of chlorine by perchlorate could also chemically alter the products of heating experiments. This may explain why so far we have only detected simple chlorinated organic molecules on Mars.

Like previous missions to Mars, Curiosity is detecting only simple organic compounds with chlorine attached. Image Credit: NASA/JPL-Caltech/MSSS

Like previous missions to Mars, Curiosity is detecting only simple organic compounds with chlorine attached. Image Credit: NASA/JPL-Caltech/MSSS

I wanted to investigate the question as to whether perchlorate is the only mineral that might have a negative influence on our search for organic compounds on Mars. I analysed a group of minerals called sulfates. They contain sulfur and oxygen in the form SO4 and include common minerals such as gypsum. When sulfates thermally break down they release sulfur dioxide and oxygen, so they have the potential to be problematic like perchlorate. However, most break down at very high temperatures (above 1000 °C), which is sufficiently high not to interfere with the release of organic molecules from samples during heating experiments. However, iron sulfates start to break down at dramatically lower temperatures. They can decompose to give off sulfur dioxide and oxygen from around 500 °C. This is around the same temperatures that large complex organic molecules might start to break down and be detectable. I was particularly interested in an iron sulfate called jarosite, as it has been detected on Mars, including recent detections by Curiosity Rover, and forms in wet acidic conditions. It’s therefore indicative of ancient wet environments that existed on Mars and may have once been inhabited by microorganisms, as similar environments on present day Earth, such as Río Tinto in Spain are a habitat for acid resistant bacteria.

I conducted fieldwork on a small island in the south of the United Kingdom called Brownsea Island. If you walk along the southern coast of Brownsea you will often see crusts of a soft yellow mineral on the short cliffs. This is jarosite, it grows here because the clay rich rocks that make up the cliff face contain the iron and sulfur mineral pyrite, pyrite reacts with water and the atmosphere to form jarosite. The geology here is a perfect case study as the rocks also contain a tough form of organic matter called lignite, a low rank of coal. I crushed the sample into a powder so that I had a mix of jarosite, clay and organic compounds. I then heated this powder at different temperatures to see if I would be able to detect the organic compounds contained in the sample. Unfortunately all I could detect was carbon dioxide, carbon monoxide, water and sulfur dioxide. The first three are compounds that you would expect to detect if organic matter was breaking down and reacting with oxygen and the sulfur dioxide indicated that the jarosite was thermally decomposing. When a sulfate breaks down we know that sulfur dioxide is paired with oxygen but when I heated this sample the oxygen wasn’t detectable. It had been consumed by reacting with organic compounds and breaking them down. From these results jarosite can now be added to the list of problematic minerals on Mars, alongside perchlorate.

Jarosite is a soft yellow mineral and can be seen growing on the clay rich cliffs of Brownsea Island, UK. As it is an iron mineral it can rust if exposed at the surface long enough in wet conditions. The orange-brown layer at the base of the cliff and the dark patches in the hand sample are rust. Image Credit: James Lewis.

Jarosite is a soft yellow mineral and can be seen growing on the clay rich cliffs of Brownsea Island, UK. As it is an iron mineral it can rust if exposed at the surface long enough in wet conditions. The orange-brown layer at the base of the cliff and the dark patches in the hand sample are rust. Image Credit: James Lewis.

Jarosite is indicative of environments that may have been habitable for life so simply avoiding it is not a satisfactory solution. Though it has a major negative influence on organic detection experiments some interpretation may still be possible. If sulfur dioxide and carbon dioxide peak at the same time in Curiosity Rover data, from a sample known to contain jarosite, it may be evidence that organic matter was present and reacting with oxygen. Unfortunately it is not always the case that a carbon dioxide peak means the presence of organic matter. Minerals known as carbonates contain carbon and oxygen in the form CO3. When carbonates thermally decompose they produce carbon dioxide. Therefore the chance of a carbonate being the source of carbon dioxide seen in Curiosity Rover data must be considered. Fortunately Curiosity has the ability to perform an assessment of the mineralogy it is adding to its heating ovens for analysis, so the presence of carbonates can be checked.

Identifying which rock units on Mars might contain abundant organic compounds would be of great use to future missions that might return samples to the Earth where a whole suite of laboratory techniques can be employed on samples without the tight space and energy constraints of a rover or lander.

My research is published online today in the journal of Astrobiology and will be free for all to read once the open access application is processed.

By James Lewis,  PhD Researcher at Imperial College London

References

Atreya, S.K., Mahaffy, P.R., and Wong, A.: Methane and related trace species on Mars: Origin, loss, implications for life, and habitability, Planetary and Space Science, 55, 358-369, doi:10.1016/j.pss.2006.02.005, 2007.

Glavin, D.P., Freissinet, C., Miller, K.E., et al.: Evidence for perchlorates and the origin of chlorinated hydrocarbons detected by SAM at the Rocknest aeolian deposit in Gale Crater, Journal of Geophysical Research, Planets Vol. 118, 1955-1973, doi:10.1002/jgre.20144, 2013.

Mahaffy, P.: Exploration of the Habitability of Mars: Development of Analytical Protocols for Measurement of Organic Carbon on the 2009 Mars Science Laboratory, Space Sci. Rev 135, 255-268, doi:10.1007/978-0-387-77516-6_18, 2008.

Ming, D.W., Archer, P.D., Glavin, D.P., et al.: Volatile and Organic Compositions of Sedimentary Rocks in Yellowknife Bay, Gale Crater, Mars. Science, 343, 1245267, doi:10.1126/science.1245267, 2013.