Lunar and Planetary Science XXIX
1316.pdf
THE CHARACTERISTICS OF LUNAR LAVA PONDS AS INDICATORS OF MAGMA TRANSPORT
MECHANISMS. R. Aileen Yingst and James W. Head, III, Brown University, Department of Geological Sciences, Box 1846, Providence, RI 02912, [email protected]
Introduction. Because of the low density and
thickness of the largely anorthositic lunar crust, magmatic rise and subsequent extrusion cannot occur
solely through the mechanism of buoyancy [1]. The
observation that this crust can be very thick in some
regions — up to 70 km in the region associated with
Mare Marginis for example [2] — suggests that a different mechanism is required to transport magma to
near-surface depths. A sequence of diapiric rise of
mantle material, stalling at the density boundary at the
base of the crust and overpressurization of partial melts
in the resulting reservoir could provide the force necessary to overcome the crustal density barrier, allowing
volcanic extrusion to occur [3]. Analysis of the morphological and volumetric characteristics of individual
eruptive lunar phases and comparison of resulting
source region volume estimates to the dimensions and
geometry of volcanic source plumes predicted by such
a model have provided evidence that supports this theory.
Method. In order to use the available surface
features to deconvolve the conditions that existed at
depth before and during lunar volcanic events, it is
necessary to determine those characteristics that comprise the statistical norm for a lunar eruptive episode.
As has been demonstrated in previous work [4-7], we
adopt a method in which isolated mare deposits (lava
ponds) are analyzed rather than the more stratigraphically and compositionally complex contiguous maria.
Those ponds which demonstrate homogeneity in morphology, albedo, age and multispectral signature where
available [e. g. 8-10] were considered best candidates
for individual eruptive phases and were thus included
in this study [e. g. 4,-7]. Isolated deposits which nevertheless showed any evidence for multiple phases,
such as flow fronts, were excluded. On the basis of
these criteria, the characteristics of 353 ponds across
the Moon were analyzed.
Morphology and geometry: The morphology of the lava ponds examined suggests that typical
lunar eruptive episodes result in relatively smooth,
featureless surfaces. There is a lack of such structures
as large domes and calderas that are normally associated with shallow magma reservoirs. The majority of
ponds occur in topographic lows such as basins and
craters. In terms of geometry, ponds tend to display
very high volumes. Previous studies [4- 7] have suggested a range of 250-850 km3 for ponds in some limb
and farside basins. Based upon our observations of
lava ponds distributed globally, the average volume for
ponds independent of location is ~450 km3, which falls
within the aforementioned range. Such volumes are
immense compared to typical historical terrestrial
eruptions [11] and are most comparable to the large
igneous provinces and flood basalts [12, 13]. This is
consistent with the magma transport model discussed
above [3], in that it predicts very large volumes of surface deposits to result from the overpressurized regime
from which magmas are driven. Additionally, if such a
volume estimate is typical of flows within the large
maria as well, it suggests that a nearside basin like
Crisium (~ 5.4 x 105 km3) [14] may potentially represent more than 1200 individual eruptive episodes.
Source region geometry. The volume of a
surface deposit can yield clues as to the size of the parent source region. For example, we can utilize a previously modelled relationship between the volume of a
source region and the volume of the associated extrareservoir material (conduit and surface deposit) of
1000:1 [15]. This relationship is shown schematically
in Figure 1. For an average surface volume of ~450
km3, as noted above, we add a reasonable conduit volume of 250 km3 [3], yielding an extra-reservoir volume
of ~700 km 3. This suggests a potential source region
volume of ~700,000 km 3, or a postulated ideal spherical source region with a diameter of ~100 km for a
~450 km 3 lava pond.
Model implications. Such estimates of potential source region sizes can be utilized to test the
predictions of magma transport models. For example,
as stated above, the model that currently best fits the
observational data [e.g. 4-7] is one in which the propagation of magma is due to the overpressurization of
deep-seated source regions stalled beneath a boundary
defined by the anorthositic crust [3]. We can thus use
these estimates of source region sizes based upon the
independently calculated average extruded deposit
volume noted above to calculate the amount of overpressurization beyond lithostatic pressure that would
be required from such a source region to produce the
associated deposit if magma transport occurred according to the above model [3]. We note that the
amount of rise for a propagated dike due to hydrostatic
considerations is
(hc ρc ) / ρm
[1,3], where h c is the associated crustal thickness, ρc is
the density of the crust and ρm is the density of the
mantle. To this expression we add
(P) / (ρm g),
the amount of rise predicted for a propagated dike due
to overpressurization of the parent source region [3],
where P is the amount of overpressure and g is the acceleration due to lunar gravity. For this expression, P is
derived from the compressibility and is calculated from
P = b (∆V/V)
Lunar and Planetary Science XXIX
1316.pdf
LUNAR LAVA PONDS: R. A. Yingst and J. W. Head, III
[15]. Here V represeents the original volume of the
source region at lithostatic equilibrium, while ∆V is the
volume of material that must be extruded after overpressurization occurs to return to lithostatic equilibrium and b is the bulk modulus. Thus
Ratio of reservoir volume to
volume of dike + mare deposit
~ 1000:1 [Blake, 1981]
(P) / (ρm g) + (hc ρc ) / ρm = H
References: [1] S. Solomon, PLSC 6, 1021, 1975.
[2] M. Zuber et al., Science 266, 1839, 1994. [3] J.
Head and L. Wilson, G&CA 56, 2155, 1992. [4] L.
Gaddis, Brown Univ. Sc.M. thesis, 1981. [5] R. A.
Yingst and J. Head, LPSC 25, 1531, 1994. [6] R. A.
Yingst and J. Head, LPSC 27, 1479, 1996. [7] R. A.
Yingst and J. Head, JGR 102, 10,909, 1997. [8] C.
Pieters, LPSC 9, 2825, 1978. [9] R. A. Yingst and J.
Head, LPSC 28, 1609, 1997. [10] J. Gillis, P. Spudis
and B. Bussey, LPSC 28, 419, 1997. [11] S. Reidel and
P. Hooper, GSA SP-239, 1989. [12] T. Tolan et al.,
GSA SP-239, 1, 1989. [13] J. Head and M. Coffin,
AGU Geophys. Mon. 100, 411, 1997. [14] J. Head et
al., G&CA, supp. 9, 43, 1978. [15] S. Blake, Nature
289, 783, 1981.
Dike driven out of reservoir
by P > Plithostatic
Crust
Dike
P = Plithostatic
Mantle
Reservoir
Figure 1. Schematic representation of relationship between intrusive
(reservoir) and extrusive (deposit + dike) volumes.
120
100
Extrusion possible
where H > T c
80
H = Tc
yields H, the total rise height of a dike propagated from
a deep-seated source region as predicted by the model
[3]. Extrusion would occur where the rise height of the
associated dike was equal to or exceeded the overlying
crustal thickness value. We use the critical value Hcrit
to represent the rise height of a dike that just reaches
the lunar surface. Thus, equating Hcrit to crustal thickness hc yields the range of dike rise height values at
which the model predicts rising dikes intersect the surface. Graphing this set of values against a range of
potential source region sizes demonstrates the range of
reservoir sizes where extrusion is possible based upon
this model (Figure 2). For this figure, the region above
the graphed line represents conditions where extrusion
to the surface is possible, and the region below the
graphed line represents conditions where no extrusion
can occur. Based upon these data, we see that the
model allows extrusion for the range of crustal thicknesses where maria already exist (negligible - 70 km
[2]) from parent source regions approximately 75-150
km in diameter. The range of reservoir diameters
yielding overpressure values that would produce observed volumes is relatively narrow. Smaller reservoirs
yield unreasonable overpressure estimates, while larger
source regions cannot provide adequate pressure to
drive a conduit to the surface. We note that the calculated source region estimate given above based upon
surface volumes (100 km) lies within this predicted
range. To a first order, then, the predictions of a
magma transport model of overpressurized source regions [3] are consistent with the observed average volumes of lunar lava ponds.
Mare Deposit
60
40
20
0
No extrusion
where H < T c
60
80
100
120 140
Reservoir diameter
160
Figure 2. Critical dike height (H) vs. reservoir size