Mapping of Martian Neutrons by HEND on Odyssey

Gamma-Ray Spectrometer suite provides synchronous measurements by all three detectors of the suite (GSH, NS and HEND). The Central Electronic Block generates a single synchronization pulse for all three, which starts the accumulation of new energy spectra of counts from each sensor and ends the accumulation of previous ones. Presently, the time interval between two successive pulses is about 20 seconds. This means that the mapping data of GRS have a temporal resolution of about 20 sec.

Mars Odyssey has a circular polar orbit with an altitude of about 400 km. With an orbital velocity of about 3 km/sec, it covers 60 km on the surface during a time of 20 sec of data acquisition. This length is the minimal scale of horizontal resolution of orbital mapping by GRS. The physical scale of surface resolution is much larger. All three instruments of Odyssey GRS have no collimators of incoming neutrons or gamma-rays. Formally speaking, any gamma-ray photon or neutron emitted from a point below the visible horizon of the planet could be detected. However, the actual angular diagram of detected neutrons is narrower than the solid angle of the visible planet. The main reason is a finite column density of the Martian atmosphere (15-25 g cm-2) along the zenith, which effectively "removes" particles from directions with a large inclination to the zenith. Also, the surface itself does not radiate neutrons isotropically into the upward semi-sphere. Neutrons at different energies have different depths of predominant generation, and they have different free paths in the subsurface. One may suppose that the surface resolutions of orbital measurements are about 300 km, which are better than the distance to the visible horizon from the orbit, but much worse than the length (60 km) of a single exposure.

We believe that surface elements with linear scale of 200-300 km provide the adequate discretization for the mapping of neutron emission of Mars. There are methods of enhanced surface resolution for orbital mapping (e.g. the method of "inverse projection"), but they are model-dependent, and we have not used them at the initial stage of data analysis.

Emission of neutrons from Mars depends on seasons of the Martian year, because during the winter atmospheric carbon dioxide deposits on the poleward surface of the planet. The thickness of the deposition depends on the season, and in the maximum it varies from one place to another from several centimeters up to about 1-2 meters [12, 13], and the production of neutrons significantly varies during the season of CO2 deposition on the surface [9]. The effect of seasonal variations of neutrons and gamma-rays provides an independent observational method of "weighing" of CO2 winter deposits in addition to direct measurements of elevation by the laser altimeter (MOLA). It is also quite useful for an independent "in-flight" calibration of the instrument by providing a "reference signal" from the layer on neutron production with well-known nuclear composition.

The study of the seasonal CO2 cycle by neutrons has been recently developed into a separate area of investigation (see [14-16]), which provides quite an important impact on the main studies of the Martian subsurface. One may determine the summer season for high-latitude regions of Mars, when the surface is CO2-free, and its nuclear emission represents the actual composition of 1-2 meters of the subsurface.

Hereafter, we present a synthetic map of the leakage flux of epithermal neutrons from the Martian surface based on the data for the summer season in the northern and southern hemispheres (Fig. 5.8). This global map does not represent any particular season on Mars because the planet usually has seasonal deposition of atmospheric carbon dioxide either in the north or south. The synthetic map of the CO2-free surface displays the permanent residual caps at both poles because they are not subject to seasonal variations.

The mapping of Mars from the orbit is influenced by the variable thickness of the atmosphere. Firstly, it has variable thickness above regions with different surface elevations [17]. Even excluding Olympus Mons, the difference of elevations across the Martian surface is about 20 km, which results in a difference in the atmospheric thickness of about 10 g cm-2along the nadir direction. One has to take these variations into account when one transforms variations of orbital data of neutron flux into variations of subsurface composition. Also, the atmosphere changes seasonally, and even mapping of the CO2-free surface of Mars is season-dependent because of that. Orbital data for each region of Mars should be associated with a set of season intervals (as AL=15°), when the thickness of the atmosphere could be considered as constant. The Ames general circulation model (GCM) [18] is thought to present quite accurate values for describing both regional and seasonal variations of the Martian atmosphere. We use this model for orbital data deconvolution below.

The most important property of the map of epithermal neutron flux of Mars is its very large dynamical range of regional flux variations (Fig. 5.8). The largest flux of epithermal neutrons is associated with the region of Solis Planum with coordinates 246-293°E, 16-53°S. The smallest fluxes are observed at two high-latitude spots in the northern and southern hemisphere with coordinates (24-28°E, 86-90°N) and (220-224°E, 78-82°S), respectively. The ratios between the maximal and minimal fluxes of epithermal neutrons in these regions are about 9.4 and 8.8, respectively.

Such a large dynamic range of regional variations (about one order of magnitude) is the main difference between the neutron albedo of Mars and that of the Moon. According to mapping data by the Lunar Prospector [11], regional variations of neutron flux of the Moon are about 15-20 %. They are thought to result from variations in the composition of major soil elements in the lunar subsurface. The surface of Mars may have quite a similar composition of major elements, but the much larger range of regional neutron variations is thought to be associated with the variable content of subsurface hydrogen, i.e. of subsurface water. Hydrogen is a very efficient moderator of high energy neutrons (see Section 5.2), and even small variations of hydrogen (about a few wt %) in the subsurface should produce significant regional variations of the leakage flux of epithermal neutrons across the Martian surface.

Fig. 5.8 Global map of neutron flux from Mars according to orbital measurements on HEND onboard Mars Odyssey. The upper map presents epithermal neutrons (signal SMD from the sensor MD) and the lower map fast neutrons (signal SSC/IN from the sensor SC/IN).

The map of epithermal neutrons displays two large poleward regions of strong depressions of the leakage flux (Fig. 5.8). According to preliminary estimations, the subsurface layer of these regions may contain tens of % of water ice by weight

[5-9]. The boundary of the southern region of neutron depression is consistent with the southern boundary of stable water ice in the subsurface [5]. Some parts of the northern depression are far southward of the northern boundary of stable water ice. A possible regional correlation has been found between the boundary of neutron depressions and the presence of polygonal polar terrains (see Fig. 7.15 in Chap. 7). Therefore, one may suggest that the northern and southern regions of neutron depressions on Mars contain very high water ice contents of several tens of wt %. More accurate evaluations of the subsurface composition of the permafrost regions are presented below, addressing several intriguing questions resulting from preliminary studies:

1) Why do northern and southern permafrost regions look so similar (see Fig. 5.8), while the surfaces of these regions are known to have very different geological origin, age, relief properties, elevation, etc.?

2) The estimated volume of water ice of several tens of wt % is much larger than the pore volume of the regolith; how has the subsurface layer with a high content of water ice been formed?

3) What is the main physical condition that determines the observed boundaries of permafrost regions? Is the visible boundary the place where the ice-rich layer extends down to the depth below the thickness of the observable neutron-producing layer, or is it the place of wet-to-dry transition of the subsurface soil?

The answers to these questions are far from complete, but we will try to address them in the following Sects. 5.5 and 5.6 to stimulate further studies and theoretical investigations.

Other surprising features on the neutron maps of Mars are the two large depressions of neutrons at moderate latitudes (Fig. 5.8). There are two opposite equatorial regions with depression of epithermal neutrons: one is the Arabia region with coordinates (0-45°E, 20°S-30°N) and another is the Memnonia region with coordinates (180-200°E, 0-25°S). The flux of neutrons in these two regions is about 3 times lower in respect to the maximum (Solis Planum). It is not as large as the depression by a factor of 9 in two polar permafrost regions, but it is still sufficiently large for regions at the equator. According to the preliminary numerical estimation (Fig. 5.3), the soil in these regions should contain about 10 wt % water to account for this depression factor, which is very difficult to explain according to the equatorial condition on the planet. This problem will also be addressed in the following section.

0 0

Post a comment