This is a topic dear to my own research interests, so I’ve been anticipating this paper for quite some time. The European Space Agency’s Mars Express probe has been orbiting Mars since December 25th, 2003 but only recently finished deploying the three radar booms for its MARSIS radar instrument. So, while we’ve been treated to gorgeous perspective photographs of the Martian surface from the HRSC stereo camera, our first glimpses of the subsurface of Mars have had to wait. But, from the first two results presented in this paper, it appears that the wait was definitely worth it.
Citation (online at CiteULike.org):
Picardi, G. and 33 others. Radar Soundings of the Subsurface of Mars. Science 310 (5756), pp. 1925-1928.
Synopsis:
If you are unfamiliar with images obtained from seismic or radar techniques, it might be best to begin by reading the explanation of subsurface radar sounding below. With that woefully incomplete explanation, hopefully the figures shown below will make a bit more sense.
Picardi and his co-authors present radargram images from three different “ground-tracks”. Each ground-track represents the ground location paths of the Mars Express orbiter during portions of its orbit where the MARSIS instrument was collecting data. The first ground track they discuss is shown at the left in part C. The ground track crosses part of the North Polar Layered Deposits where a polar ice cap is suspected and has been observed in images as well as by other probes and their instruments. Part A of the figure on the left is the actual radargram from the MARSIS instrument. Going from left to right, their is a single bright surface reflection that splits into two at about the middle of the image. The top reflector rises while the bottom remains relatively flat. Part B shows the modeled results of what is expected given surface topography. The bottom reflector, not expected from surface reflections alone, is a “basal reflection”. Picardi and his co-authors note that this radargram appears to be showing exactly what we would expect: there is a large cap of mostly ice situated atop a relatively flat northern plain. They are able to say that the cap is ice because the strength of the radar signal diminishes very little between the top and bottom reflectors. This very small amount of “signal attenuation” is indicative of cold water ice.
The radargrams from ground tracks two and three are shown in the parts A and B of the figure to the right. Part C is a model of the expected signal for the ground track in part B. These two ground tracks are separated by about 30 km and are roughly parallel. The top bright white line in each radargram shows the surface reflection itself, while the arc-shaped reflections in the center of each image indicate the presence of a buried arc of material at some distance away from being directly below the orbiter itself. (See the image in the General Explanations section, below). The second ground track, shown in part B also exhibits a flat reflection beneath the surface. The authors suggest that this may, in fact, be a reflection from the top of a flat layer of ice that partially fills the buried crater. However, they note that the absence of the flat reflection signal in the first ground track weakens this hypothesis.
Whether or not there is an ice-deposit within the crater itself, the authors make a strong case that the arcs are due to a buried crater not seen in either the laser-topography data or visual images. They also note that the arcs can not be fully explained by only a single crater wall reflection from each direction, but rather there are two such reflections, which may again be indicative of a partially-filled crater (see the image below).
They plot the likely location of these crater rim reflectors on a map along with the two ground tracks shown above. Notice that the western arcs are roughly coincident, while the eastern arcs differ slightly between the two ground tracks. Assuming that the outer set of arcs corresponds to the rim of the crater itself, the crater is approximately 250 km across. A crater this size was very likely caused by an impact rather than volcanism. Also, such an impact would perhaps have had important effects on the global Martian environment, and locating these buried craters that may be quite common in the Northern Plains will be very important in understanding Mars’ geologic history.
Context:
These two results are the especially exciting considering they came from data collected within the first few months of what may be a many-year long MARSIS lifetime. Continued collection and analysis will allow further detection of buried features, perhaps even of large frozen water deposits like that potentially seen in this buried crater. The MARSIS probe has opened up an entirely new realm of investigation on Mars, no longer are we limited to its surface. Short of having humans on the surface drilling, we have no way of getting glimpses of the subsurface composition of this unexplored world other than via radar. The NASA probe currently on its way to Mars, the Mars Reconnaissance Orbiter carries a similar instrument, so these radar images may become much more familiar to us all over the coming years.
General Explanations:
The MARSIS radar collects data for approximately 26 minutes of each 6.7 hour orbit of the Mars Express probe, continually pinging Mars with radar waves between 1.3 and 5.5 MHz. These low-frequency radar pulses are able to penetrate the Martian crust and travel several kilometers down before being absorbed. Each time the radar waves encounter a sudden change in electric properties of the crust (for example, between air and rock, ice and rock, or even different rock types) a portion of the radar signal is reflected back to the receiver on the Mars Express probe. After each radar ping, MARSIS “listens” for the radar echoes from the planet far below.
However, the radar signal illuminates an area approximately 30×30 km on the surface, so identifying exactly where an echo came from on or below the surface can be difficult. To overcome this difficulty, the MARSIS probe performs synthetic aperture radar processing, and scientists back on earth simulate the effects of surface clutter on computers. Thus, anything not seen in those simulations can be assumed to be beneath the surface.
When viewing radar images, the most important thing to remember is that what you see is not a picture of what’s immediately below the spacecraft. The picture is instead constructed by joining successive columnar plots of echoed signal strength (shown in the plots above as greyscale intensity) versus the time since the radar ping (along the vertical axis, time increases in the downward direction). This means that the arcs seen in the second figure are not showing the vertical profile of a crater, but rather show the travel time between the spacecraft and the crater wall. The “deepest” arcs indicate the portion of the crater furthest from the orbit of the spacecraft. The authors illustrate this effect in another figure from their paper, shown at the right. In this figure, the path of the Mars Express is shown by the arrow, while the bright white reflections seen in the radargrams in the two figures above are depicted as colored lines on the plot. The physical location along the ground of those lines is shown in the perspective view cartoon.
Radar and seismic imaging is a powerful tool, but it is not capable of producing pictures like those of the buried raptor in Jurassic Park. That’s not because ground penetrating radar is an immature technology, but simply because of the physical reality of what seismic and radar imaging can produce. (image credit: Universal Studios, captured DVD still).
