11 - 5

GDS

HEGGY ET AL.: MARTIAN GEOELECTRICAL MODELS

[Hargraves et al., 1977] and 2% of magnetite, thus using a

mean value of 15% of iron oxide concentration in the

Martian surface dust layer. Its relative low dielectric value

is due the high porosity around 50% (even if it is rich in iron

oxides). Samples representing the dust layer have a grain

size of 50 mm, which is the observed value for hematite

grain size at the Martian surface [Bell and Morris, 1998].

For the samples representing the subsurface material, we

considered a larger grain size (200 to 400 mm). This

parameter is very important specially in measuring the

sample magnetic permeability for iron oxide-rich materials,

and it also controls the samples porosity. Another common

layer for the three volcanic models is the water-saturated

layer denoted by ``wet basalt'' in Figures 1, 2, and 3, which

was simulated using water saturated basalt powder.

[15] The geoelectrical properties of the Hadriarca Patera

volcano site are presented in Table 1. We simulated the

second layer (eroded basalt) using a rock-machined pellet of

Djiboutian basalt, which presents very similar chemical and

physical properties to the rock analysis provided by the

Viking and Pathfinder landers [Paillou et al., 2001]. To

simulate experimentally the third layer constituting the

ground ice, we mixed a basalt powder to water and we

compacted it to reach the lithospheric pressure at the

corresponding depth in the geological model. The mixture

was then put in a cold room down to the 210°K temperature.

Special precautions were taken to ensure that samples

(volcanic and ice mixed minerals) were free of moisture.

[16] For the outwash plains geoelectrical model presented

in Table 2, we simulated experimentally the fluvial sedi-

mentary layer, by measuring the permittivity of a powder of

basalt mixed with 25% mass percentage of aragonite and

dolomite at a porosity of 40%. For the lava and the ground

ice layers, we used respectively a compacted dry basalt

Figure 4. Top: MOC image of the Melas Chasma region

powder with a porosity of 35% and a basalt rock machined

(9 S, 77.5 W) (MOC image M08-04367). The arrow

pellets with a lower porosity of 25% with ice inside the

indicates what could represent a dry mudstone eroded layer.

pores. We can clearly note the difference in their dielectric

Bottom: the suggested model for this type of layered terrain.

constant, which is mainly due to the difference in porosity

between the two samples (which corresponds to a different

lithospheric pressure in the geological profile). At a greater

depth, we have a higher compaction leading to a lower

[13] Tables 1, 2, 3, and 4 present the geoelectrical profiles

porosity and thus a higher dielectric constant of the material.

for the geological models shown in Figures 1, 2, 3, and 4,

[17] In the ejecta deposits model shown in Table 3, we

respectively. In those models we assumed that the layers are

used a low compacted basalt powder of 300 mm grain size

homogenous and parallel according to the observed stratig-

mixed to the powder constituting the dust layer with 10% of

raphy on the exposed wall rock of Valles Marineris [McE-

ice. We mixed basalt and silicate to simulate the regolith

wen et al., 1999]. The interfaces between layers have a step

layer. For the eroded basalt layer, we measured the permit-

periodic roughness function with maximum amplitude of 1

tivity of basalt mixed with 5% of hematite. As representa-

m, which means that the shallow interfaces are relatively

tive sample of the bottom layer in this geological profile, we

smooth compared to the wavelength inside the materials for

used highly compacted basalt (porosity <20%). The dust

2 MHz radars. We introduced also a fuzzy level at each

layer and the wet saturated layer have been treated similarly

interface to take into account a short material transition

as in the previous models.

gradient between each layer and possible unfrozen water

[18] The layered deposits model presented in Table 4

concentration gradient [Anderson and Morgenstern, 1973].

[14] According to the surface chemical analysis of the

corresponds to a quite different geological context. In this

Viking Landers and the Pathfinder mission suggest that the

model, we did not introduce any ferromagnetic materials,

dust layer that covers the Martian surface can be assumed to

(except at the bottom basalt bedrock) and we mainly used

be chemically homogenous [Reider et al., 1997] for the

dry powder of kaolinite mixed to miner amount of materials

major part of the planet. Thus we considered for the three

described in the dust layer to estimate the permittivity of the

models of Hadriarca Patera, ejecta deposits and outwash

first layer. The second layer was characterized with a

plains, the presence of a thin layer of dust (10 m). To

gypsum compacted pellet, and we used an aragonite com-

simulate this dust layer, we mixed a dry basalt powder with

pacted pellet to simulate the possible presence of a carbo-

a mass percentage of 7% of hematite, 7% of maghemite

nate layer in the Martian subsurface [Fonti et al., 2001].