Magma Ascent Rates

1
Magma Ascent Rates

• Malcolm J Rutherford
• Dept. of Geological Sciences
• Brown University
• Providence RI
• (Presentation for MSA short Course Dec. 13,
  2008)

2
Outline

• 1. Why study magma ascent rates?
• - Better understanding of subsurface magmas
  and processes.
• - Prediction of eruption style changes,
  e.g., explosive vs. effusive.
• 2. How can we determine ascent rates?
• - Melt degassing produced by ascent-induced
  decompression.
• - Crystallization driven by decompression
  and degassing.
• - Crystal-melt reactions
  .
• - Seismicity changes associated with magma
  ascent.
• - Theoretical modeling of magma ascent.
• 3. Caveats and problems involved?
• - Determining where magma ascent began.
• - Determining the exact nature of reactions
  involved.
• - Involvement of magma convection and
  mixing.
• - The size of the conduit.

3
Amph
Opx
4 km -
Phase equilibria of 2004-06 MSH dacite magma
showing amphibole stability limit inset Shows
internal zoning in the amphibole and Opx that
requires two cycles of convection from 300 to
120 MPa prior to final ascent (after Rutherford
and Devine, 2008).
4
Using diffusion data for water, Humphries et al.,
(2008) calculate time required to develop
measured profile in tubular melt
channels. Assuming 4.6 wt H2O initially, L 5
km, C.R. 25 m, v 37- 49 m/s for MSH May 18,
1980. Method similar to that of Anderson (1996)
Liu et al., 2007.
5
Conduit Radius (effective) and Magma ascent
velocity changes derived from magma mass eruption
rates at Mount St. Helens in 1980-82 eruptions
assuming single phase flow (after Geschwind and
Rutherford, 1995). - Ave. Ascent rate for 1980
explosive 3 m/s for 8 km (C.R. 33 m). -
Theoretical modeling V 15-20 m/s (C.R.
17-33m) Papale Dobran(1994). - Humphries et
al., (2008) V 37- 49 m/s. Conduit radius
and depth both critical in determining rate of
ascent.
6
Ascent rates (m/s)
I 0.045
I 0.022
I 0.015
I 0.01
Ascent rates based on crystallization induced by
decompression in MSH 1980-86 dacite magma
(after Geschwind and Rutherford, 1995). Assume 8
km starting depth and conduit radius from mass
eruption rate data (Swanson et al., 1987).
7
SH308 Amph
Phase equilibria of MSH dacite magma showing
amphibole stability limit inset shows amphibole
from a 2005 sample with a reaction rim (after
Rutherford and Devine, 2008).
8
Amphibole reaction rim-width growth in dacitic
and andesitic magmas as a function of
decompression time (constant rate decompression)
from experiments at 830-900 oC (after Rutherford
and Hill, 1993 Rutherford and Devine, 2003).
9
Images showing reaction rims on amphibole
phenocryst in the Black Butte (CA) dacite also
illustrated is the contrast in texture of the
phenocrysts relative to the lineated matrix.
After
McCanta et al., (2007).

• - Rim widths (34 ?m) in each of 4 lobes of BB
  lava dome erupted at 890oC, yield an ave.
• magma ascent rate of 0.004-0.006 m/s for ascent
  from 200 MPa (8 km) to the surface.
• Single stage ascent and no mixing indicated
  for these magmas.

10
840oC
New amphibole reaction rim growth rate study show
how the rate varies with P at a 840oC (Brown and
Gardner, 2006).
11
Temperature affect of decompression-induced
crystallization in MSH 1980-82 dacite
Modified after Blundy et al., (2006)
12
A
B
F wt
F-rich amphibole rims develop in very slow
ascending 2004-6 Mount St. Helens dacite magma
preventing rim growth (DeJesus and Rutherford,
2008)
13
- Zoning developed in Mt Unzen Ti-magnetite.
(Nakamura, 1995). Profile reflects time
following magma mixing, and yields a minimum
ascent time assuming that mixing accompanied
onset of ascent. - Ascent of Mt. Unzen mixed
magma from 7 km was at a min. rate of 0.003-0.007
m/s (11-30 da) based on 20 ?m profile.
14
1996-2002 Montserrat TiO2 profiles for natural
and experimental Ti-magnetite (after Devine et
al., 2003). 830-860oC andesitic magma reached
surface from 5 km depth 20 days after basaltic
andesite influx.
15
Real Time Seismic Amplitude counts build-up at
site near Mount St. Helens in 1986 along with
focal depth shows magma rising from 1.4 km to
surface over 12 hours this corresponds to an ave
magma ascent rate of 0.32 m/s (Endo et al., 1996)

16
Formation at 7-15 km
(1)
(2)
Xenolith-melt reactions. Two scales of
reaction in olivine-rich xenoliths carried up in
basalt at La Palma, Canary Islands. (1) Long
diffusion profiles in Olv 8-110 yrs storage at
7-15 km (2) Melt bearing fractures in Olv have
zoning that indicates 0-4 days origin and ascent
rates of gt 0.06 m/s (Klugel et al., 1998)
assuming cracks began with ascent.
17
Garnet-melt reaction in kimberlite (Canil and
Fedortchouk, 1999). Garnet dissolution features
(25?m) interpreted using experimental data yield
exposure times minimum ascent rates depending on
depth where garnet is exposed and T.
18
Water loss profiles in olivine in garnet
peridotite xenolith carried in alk basalt
(Demouchy et al., 2006 Peslier and Luhr, 2006).
Assuming 40 and 60-70 km depth for inclusion
enclosure in basalt respectively, initial water
300 ppm, and 1200oC, D et al., calculate 6.3 m/s
ascent P L ascent rate 0.2 - 0.5m/s.
19
Kimberlite Magma Ascent Rates from theoretical
flow modeling

• Calculation is for buoyancy-driven dike flow
  (Stage 2), does not consider effect of gas
  expansion that would be particularly important in
  Stage 3 (1- 0 km).
• Calculation agrees well with estimates from
  garnet dissolution and with xenolith transport
  requirements (Spera, 1984).

20
(No Transcript)
21
(No Transcript)
22
Dissolved volatiles in Min2 shoshonite and F.R
latite olv- and cpx-hosted melt inclusions (FTIR)
after Mangiacapra et al., 2008, GRL.
23
Lobe 1, SHM-22, lt 50um Slope
0.2260, intercept -12.45. 50-600 um, slope
0.0153, intercept -22.53
Plagioclase Phenocrysts and microlites in Black
Butte CA dacite. CSD for the two phases of
crystallization gives growth rate
24
(No Transcript)
25
(No Transcript)
26
(No Transcript)
27
Objectives

• 1. Why study magma ascent rates?
• - Better general understanding of subsurface
  volcanic processes.
• - Prediction of eruption style, e.g.,
  explosive vs. effusive.
• How can we determine ascent rates?
• - Magma degassing rates from melt phase.
• - Crystallization driven by decompression
  and degassing.
• - Crystal-melt reactions
  .
• - Seismicity associated with magma ascent.
• - Zoning developed in crystals by
  decompression.
• - Theoretical modeling of magma ascent.
• 3. Caveats and problems involved?
• - determining where magma ascent began, the
  nature of reaction observed, the parameters
  controlling the reaction, involvement of magma
  convection and mixing, the size of the conduit.

28
Magma storage zone at Mount Pinatubo in 1991
based on seismicity and petrological phase
equilibria of the phenocryst - melt assemblage
(after Pallister et. al., 1996 and Hammer, 2003).