Geological and physical observations and constraints are applied to the development of a model of the ascent and emplacement of basaltic magma on the earth and moon. Mathematical models of the nature and motion of gas/liquid mixtures are developed and show that gas exsolution from terrestrial and lunar magmas commonly only occurs at shallow depths (less than 2 km); thus the ascent of bubble‐free magma at depth can be treated separately from the complex motions caused by gas exsolution near the surface. Magma ascent is related to dike or conduit width; a lower limit to width is determined by the presence of a finite magma yield strength or by excessive magma cooling effects related to magma viscosity. For terrestrial basalts with negligible yield strengths and viscosities greater than 102 Pa s, widths in the range 0.2–0.6 m are needed to allow eruptions from between depths of 0.5–20 km. Fissure widths of about 4 m would be needed to account for the output rates estimated for the Columbia River flood basalt eruptions. As the magma nears the surface, bubble coalescence will tend to occur, leading to intermittent explosive strombolian‐style activity. For commonly occurring lunar and terrestrial basalts the magma rise speed must be greater than 0.5–1 m/s if strombolian activity is to be avoided and relatively steady fire fountaining is to take place. Terrestrial fire fountain heights are dictated by the vertical velocity of the magma/gas dispersion emerging through the vent, increasing with increasing magma gas content and mass eruption rate, and decreasing with increasing magma viscosity. Terrestrial fire fountain heights up to 500 m imply the release of up to 0.4 wt % water from the magma, corresponding to initial water contents up to 0.6 wt %. The presence of extremely long lava flows and sinuous rules on the moon has often been cited as evidence for very high extrusion rates and thus a basic difference between terrestrial and lunar magmas and crustal environments. However, the differences between terrestrial and lunar magma rheologies and crustal environments do not lead to gross differences between the effusion rates expected on the two planetary bodies, for similar‐sized conduits or fissures. Thus the presence of these features implies only that tectonic and other forces associated with the onset of some lunar eruptions were such as to allow wide fissures or conduits to form. The surface widths of elongate fissure vents need be no wider than 10 m to allow mass eruption rates up to 10 times larger than those proposed for terrestrial flood basalt eruptions; 25‐m widths would allow rates 100 times larger. It therefore appears unlikely that source vents on the moon with widths greater than a few tens of meters represent the true size of the unmodified vent. The main volatile released from lunar magmas was probably carbon monoxide, released in amounts proportionally less than terrestrial magmas by more than an order of magnitude. However, decompression to the near‐zero ambient lunar atmospheric pressure causes much g...

No abstract

Plinian air-fall deposits and ignimbrites are the principal products of explosive eruptions of hgh viscosity magma. In this paper, the flow of gas/ pyroclast dispersions and high viscosity magma through various magma chamber/conduit/vent geometries is considered. It is argued that after the first few minutes of an eruption magma fragmentation occurs at a shallow depth within the conduit system. Gas pressures at the fragmentation level are related to exsolved gas contents by consideration of the exsolution mechanism.The sizes of blocks found near vents imply that gas velocities of 200 to 600 m s-' commonly occur. These velocities are greater than the effective speed of sound in an erupting mixture (90-200ms-') and the transition from subsonic to supersonic flow is identified as occurring at the depth at which the conduit has its minimum diameter. The range of values of this minimum diameter (-5 to -100 m) is estimated from observed and theoretically deduced mass-eruption rates.The energy and continuity equations are solved, taking account of friction effects, for numerous geometries during the evolution, by wall erosion, of a conduit. Conduit erosion ceases, near the surface, when an exit pressure of one atmosphere is reached. Eruption velocities are found to depend strongly on exsolved magma gas content and weakly on radius of conduit and friction effects. Assuming water as the main volatile phase, velocities of 400-600 m s-l for plinian events imply magma water contents of 3-6 per cent by weight, Three scenarios are presented of eruptions in which: (1) conduit radius increases but gas content remains constant; (2) conduit radius increases and gas content decreases with time; and (3) conduit radius remains fured and gas content decreases. These models demonstrate that the reverse grading 118 L. Wilson, R. S. J. Sparks and G. P. L. Walker commonly observed in plinian air-fall deposits is primarily a consequence of conduit erosion, which always results in increasing eruption intensity and eruption column height with time. The models also show that a decrease in gas content as deeper levels in a magma chamber are tapped or an increasing vent radius as conduit walls are eroded leads to the prediction of a progression from air-fall activity through ignimbrite formation to cessation of eruption and caldera collapse.