Jökull

Ataaseq assigiiaat ilaat

Jökull - 01.01.2019, Qupperneq 98

Jökull - 01.01.2019, Qupperneq 98
Eyjafjallajökull ankaramites, South Iceland The diffusion modelling gives diffusion time scales in the range 9–37 days (Figure 6). This time frame does not reflect the speed of the ankaramite magma ascent directly, because of the unknown olivine residence time in the lava and the fact that the timing of the magmatic recharge (forming the reverse zoned high-Fo rims) does not necessarily coincide with the beginning of the magma ascent. The mag- matic recharge to the crystal mush did not necessar- ily act as the eruption trigger and may have occurred considerably earlier than agitation and upheaval of the olivine (and clinopyroxene) crystals (see Rae et al., 2016). Nevertheless, as the unknown time of diffusion at the surface and the potential time-delay between magmatic recharge and crystal lift both lengthen the diffusion time, our 9–37-day estimate for diffusive re- equilibration can be viewed as a likely maximum time that elapsed from crystal mush disaggregation to erup- tion. The 9–37-day range may reflect varying olivine residence times in a magma ascent conduit or surfi- cial lava-feeding system after the magmatic recharge event, or be due to olivine sectioning and anisotropy effects (Shea et al., 2015). Moderate-pressure co-crystallization of primitive clinopyroxene and olivine The Brattaskjól and Hvammsmúli ankaramites have abundant olivine and clinopyroxene macrocrysts and only minor plagioclase macrocrysts, Mg#cpx of clinopyroxene being as high as 90. This and the compositional variation in clinopyroxene cores sug- gest olivine- and clinopyroxene-dominated fractiona- tion and a late arrival of plagioclase on the liquidus. Specifically, Ca and Al enrichment with decreas- ing magnesium number in clinopyroxene grains with Mg#cpx 84.5–90.0 suggests fractionation of olivine and clinopyroxene in approximately equal amounts, and the decrease in Ca and Al, which indicates si- multaneous plagioclase fractionation from the host melt, is only seen in clinopyroxene grains with Mg#cpx<84.5 (Figure 4c and d). Variation in crystal- lization conditions (T and P) or disequilibrium crys- tallization processes are unlikely to explain this com- positional variation because of the unchanged Jd con- tent in clinopyroxene cores (Figure 4e), which should be dependent on pressure, temperature and crystal- lization rate (e.g., Mollo et al., 2010; Hammer et al., 2016; Welsch et al., 2016; Ubide et al., 2019). More- over, rapid disequilibrium crystallization is expected to produce anti-correlation between Ca and Al (Mollo et al. 2010), contrary to what is seen in our clinopy- roxene data (Figures 4c and d). MORB melts typically crystallize in the sequence of olivine (±spinel) –> olivine + plagioclase –> olivine + plagioclase + clinopyroxene, leaving be- hind troctolitic or gabbroic cumulates, not wehrlite cumulates, such as in the case of the Brattaskjól and Hvammsmúli ankaramites. In addition, due to the late arrival of clinopyroxene as a crystallizing phase, the most primitive clinopyroxene produced in MORB crystallization experiments typically has Mg#cpx of ∼83 (Grove and Bryan, 1983; Tormey et al., 1987; Yang et al., 1996), not Mg#cpx 90 like in Bratta- skjól ankaramite. Geochemical trends indicative of clinopyroxene fractionation before plagioclase have been noted from lavas in SEVZ and other parts of Iceland (Furman et al., 1991; Thy, 1991a; Maclen- nan et al., 2001; Mattsson and Oskarsson, 2005), and some Iceland lavas do host high-Mg# clinopyroxene (up to Mg#cpx 92 in Borgarhraun lava, Winpenny and Maclennan 2011). In this regard, our findings are in line with earlier studies. Anyhow, clinopyroxene frac- tionation from primitive SEVZ lavas has usually been regarded as indicative of high (>8 kbar) crystallization pressures near Moho or at upper mantle depths (Fur- man et al., 1991; Thy, 1991a, b; Mattsson and Oskars- son, 2005), as the stability of clinopyroxene increases with pressure (Presnall et al., 1978, 2002; Stolper, 1980). The calculated 3.0±1.4 kbar crystallization pressures for the primitive Brattaskjól and Hvamms- múli clinopyroxene macrocrysts (up to Mg#cpx 87) contradicts this contention and suggest that the mildly alkaline SEVZ magmas crystallized clinopyroxene and olivine before plagioclase at depths as shallow as in the mid-crust. This inference is supported by, and is in line with, recent experimental work that shows that "enriched end-member" (Shorttle and Maclen- nan, 2011) Iceland tholeiite melt crystallized clinopy- roxene before plagioclase at 3 kbar (Neave et al., 2019b). Elevated H2O in primitive SEVZ magmas (Thy, 1991a; Moune et al., 2012) and low amounts of JÖKULL No. 69, 2019 97
Qupperneq 1
Qupperneq 2
Qupperneq 3
Qupperneq 4
Qupperneq 5
Qupperneq 6
Qupperneq 7
Qupperneq 8
Qupperneq 9
Qupperneq 10
Qupperneq 11
Qupperneq 12
Qupperneq 13
Qupperneq 14
Qupperneq 15
Qupperneq 16
Qupperneq 17
Qupperneq 18
Qupperneq 19
Qupperneq 20
Qupperneq 21
Qupperneq 22
Qupperneq 23
Qupperneq 24
Qupperneq 25
Qupperneq 26
Qupperneq 27
Qupperneq 28
Qupperneq 29
Qupperneq 30
Qupperneq 31
Qupperneq 32
Qupperneq 33
Qupperneq 34
Qupperneq 35
Qupperneq 36
Qupperneq 37
Qupperneq 38
Qupperneq 39
Qupperneq 40
Qupperneq 41
Qupperneq 42
Qupperneq 43
Qupperneq 44
Qupperneq 45
Qupperneq 46
Qupperneq 47
Qupperneq 48
Qupperneq 49
Qupperneq 50
Qupperneq 51
Qupperneq 52
Qupperneq 53
Qupperneq 54
Qupperneq 55
Qupperneq 56
Qupperneq 57
Qupperneq 58
Qupperneq 59
Qupperneq 60
Qupperneq 61
Qupperneq 62
Qupperneq 63
Qupperneq 64
Qupperneq 65
Qupperneq 66
Qupperneq 67
Qupperneq 68
Qupperneq 69
Qupperneq 70
Qupperneq 71
Qupperneq 72
Qupperneq 73
Qupperneq 74
Qupperneq 75
Qupperneq 76
Qupperneq 77
Qupperneq 78
Qupperneq 79
Qupperneq 80
Qupperneq 81
Qupperneq 82
Qupperneq 83
Qupperneq 84
Qupperneq 85
Qupperneq 86
Qupperneq 87
Qupperneq 88
Qupperneq 89
Qupperneq 90
Qupperneq 91
Qupperneq 92
Qupperneq 93
Qupperneq 94
Qupperneq 95
Qupperneq 96
Qupperneq 97
Qupperneq 98
Qupperneq 99
Qupperneq 100
Qupperneq 101
Qupperneq 102
Qupperneq 103
Qupperneq 104
Qupperneq 105
Qupperneq 106
Qupperneq 107
Qupperneq 108
Qupperneq 109
Qupperneq 110
Qupperneq 111
Qupperneq 112
Qupperneq 113
Qupperneq 114
Qupperneq 115
Qupperneq 116
Qupperneq 117
Qupperneq 118
Qupperneq 119
Qupperneq 120
Qupperneq 121
Qupperneq 122
Qupperneq 123
Qupperneq 124
Qupperneq 125
Qupperneq 126
Qupperneq 127
Qupperneq 128
Qupperneq 129
Qupperneq 130
Qupperneq 131
Qupperneq 132
Qupperneq 133
Qupperneq 134
Qupperneq 135
Qupperneq 136
Qupperneq 137
Qupperneq 138
Qupperneq 139
Qupperneq 140
Qupperneq 141
Qupperneq 142
Qupperneq 143
Qupperneq 144
Qupperneq 145
Qupperneq 146
Qupperneq 147
Qupperneq 148
Qupperneq 149
Qupperneq 150
Qupperneq 151
Qupperneq 152
Qupperneq 153
Qupperneq 154
Qupperneq 155
Qupperneq 156
Qupperneq 157
Qupperneq 158
Qupperneq 159
Qupperneq 160
Qupperneq 161

x

Jökull

Direct Links

Hvis du vil linke til denne avis/magasin, skal du bruge disse links:

Link til denne avis/magasin: Jökull
https://timarit.is/publication/1155

Link til dette eksemplar:

Link til denne side:

Link til denne artikel:

Venligst ikke link direkte til billeder eller PDfs på Timarit.is, da sådanne webadresser kan ændres uden advarsel. Brug venligst de angivne webadresser for at linke til sitet.