Generation, ascent and eruption of magma on the Moon:new insights into source depths, magma supply, intrusions and effusive/explosive eruptions (Part 1: Theory)

Wilson, Lionel and Head, James W. (2017) Generation, ascent and eruption of magma on the Moon:new insights into source depths, magma supply, intrusions and effusive/explosive eruptions (Part 1: Theory). Icarus, 283. pp. 146-175. ISSN 0019-1035

PDF (ICARUS-lunar-part-1)
ICARUS_lunar_part_1.pdf - Accepted Version
Available under License Creative Commons Attribution-NonCommercial-NoDerivs.

Download (1MB)


We model the ascent and eruption of lunar mare basalt magmas with new data on crustal thickness and density (GRAIL), magma properties, and surface topography, morphology and structure (Lunar Reconnaissance Orbiter). GRAIL recently measured the broad spatial variation of the bulk density structure of the crust of the Moon. Comparing this with the densities of lunar basaltic and picritic magmas shows that essentially all lunar magmas were negatively buoyant everywhere within the lunar crust. Thus positive excess pressures must have been present in melts at or below the crust–mantle interface to enable them to erupt. The source of such excess pressures is clear: melt in any region experiencing partial melting or containing accumulated melt, behaves as though an excess pressure is present at the top of the melt column if the melt is positively buoyant relative to the host rocks and forms a continuously interconnected network. The latter means that, in partial melt regions, probably at least a few percent melting must have taken place. Petrologic evidence suggests that both mare basalts and picritic glasses may have been derived from polybaric melting of source rocks in regions extending vertically for at least a few tens of km. This is not surprising: the vertical extent of a region containing inter-connected partial melt produced by pressure-release melting is approximately inversely proportional to the acceleration due to gravity. Translating the ∼25 km vertical extent of melting in a rising mantle diapir on Earth to the Moon then implies that melting could have taken place over a vertical extent of up to 150 km. If convection were absent, melting could have occurred throughout any region in which heat from radioisotope decay was accumulating; in the extreme this could have been most of the mantle. The maximum excess pressure that can be reached in a magma body depends on its environment. If melt percolates upward from a partial melt zone and accumulates as a magma reservoir, either at the density trap at the base of the crust or at the rheological trap at the base of the elastic lithosphere, the excess pressure at the top of the magma body will exert an elastic stress on the overlying rocks. This will eventually cause them to fail in tension when the excess pressure has risen to close to twice the tensile strength of the host rocks, perhaps up to ∼10 MPa, allowing a dike to propagate upward from this point. If partial melting occurs in a large region deep in the mantle, however, connections between melt pockets and veins may not occur until a finite amount, probably a few percent, of melting has occurred. When interconnection does occur, the excess pressure at the top of the partial melt zone will rise abruptly to a high value, again initiating a brittle fracture, i.e. a dike. That sudden excess pressure is proportional to the vertical extent of the melt zone, the difference in density between the host rocks and the melt, and the acceleration due to gravity, and could readily be ∼100 MPa, vastly greater than the value needed to initiate a dike. We therefore explored excess pressures in the range ∼10 to ∼100 MPa. If eruptions take place through dikes extending upward from the base of the crust, the mantle magma pressure at the point where the dike is initiated must exceed the pressure due to the weight of the magmatic liquid column. This means that on the nearside the excess pressure must be at least ∼19 ± 9 MPa and on the farside must be ∼29 ± 15 MPa. If the top of the magma body feeding an erupting dike is a little way below the base of the crust, slightly smaller excess pressures are needed because the magma is positively buoyant in the part of the dike within the upper mantle. Even the smallest of these excess pressures is greater than the ∼10 MPa likely maximum value in a magma reservoir at the base of the crust or elastic lithosphere, but the values are easily met by the excess pressures in extensive partial melt zones deeper within the mantle. Thus magma accumulations at the base of the crust would have been able to intrude dikes part-way through the crust, but not able to feed eruptions to the surface; in order to be erupted, magma must have been extracted from deeper mantle sources, consistent with petrologic evidence. Buoyant dikes growing upward from deep mantle sources of partial melt can disconnect from their source regions and travel through the mantle as isolated bodies of melt that encounter and penetrate the crust–mantle density boundary. They adjust their lengths and internal pressure excesses so that the stress intensity at the lower tip is zero. The potential total vertical extent of the resulting melt body depends on the vertical extent of the source region from which it grew. For small source extents, the upper tip of the resulting dike crossing the crust–mantle boundary cannot reach the surface anywhere on the Moon and therefore can only form a dike intrusion; for larger source extents, the dike can reach the surface and erupt on the nearside but still cannot reach the surface on the farside; for even larger source extents, eruptions could occur on both the nearside and the farside. The paucity of farside eruptions therefore implies a restricted range of vertical extents of partial melt source region sizes, between ∼16 and ∼36 km. When eruptions can occur, the available pressure in excess of what is needed to support a static magma column to the surface gives the pressure gradient driving magma flow. The resulting typical turbulent magma rise speeds are ∼10 to a few tens of m s−1, dike widths are of order 100 m, and eruption rates from 1 to 10 km long fissure vents are of order 105 to 106 m3 s−1. Volume fluxes in lunar eruptions derived from lava flow thicknesses and surface slopes or rille lengths and depths are found to be of order 105 to 106 m3 s−1 for volume-limited lava flows and >104 to 105 m3 s−1 for sinuous rilles, with dikes widths of ∼50 m. The lower end of the volume flux range for sinuous rilles corresponds to magma rise speeds approaching the limit set by the fact that excessive cooling would occur during flow up a 30 km long dike kept open by a very low excess pressure. These eruptions were thus probably fed by partial melt zones deep in the mantle. Longer eruption durations, rather than any subtle topographic slope effects, appear to be the key to the ability of these flows to erode sinuous rille channels. We conclude that: (1) essentially all lunar magmas were negatively buoyant everywhere within the crust; (2) positive excess pressures of at least 20–30 MPa must have been present in mantle melts at or below the crust–mantle interface to drive magmas to the surface; (3) such pressures are easily produced in zones of partial melting by pressure-release during mantle convection or simple heat accumulation from radioisotopes; (4) magma volume fluxes available from dikes forming at the tops of partial melt zones are consistent with the 105 to 106 m3 s−1 volume fluxes implied by earlier analyses of surface flows; (5) eruptions producing thermally-eroded sinuous rille channels involved somewhat smaller volume fluxes of magma where the supply rate may be limited by the rate of extraction of melt percolating through partial melt zones.

Item Type:
Journal Article
Journal or Publication Title:
Additional Information:
This is the author’s version of a work that was accepted for publication in Icarus. Changes resulting from the publishing process, such as peer review, editing, corrections, structural formatting, and other quality control mechanisms may not be reflected in this document. Changes may have been made to this work since it was submitted for publication. A definitive version was subsequently published in Icarus, 283, 2017 DOI: 10.1016/j.icarus.2015.12.039
Uncontrolled Keywords:
ID Code:
Deposited By:
Deposited On:
01 Mar 2016 11:58
Last Modified:
08 Apr 2020 03:07