Deep upper mantle phases missing from MAGEMin 1.3.2
Closed this issue · 2 comments
Issue
Hi MAGEMin team,
I am calculating stable solutions for a dry upper mantle between 1–28 GPa and 773–2773 K. MAGEMin is apparently missing some important phase transitions in the lower upper mantle (e.g., wadsleyite, ringwoodite, perovskites, majorite ?).
I see that all of the necessary end members are included in the source code from tc-ds634, but I believe the EOS for many lower upper mantle minerals are missing in the current version of the program (MAGEMin 1.3.2 [06/06/2023]). Up to about 10 GPa the mineral assemblages are consistent with Perple_X, but above 10 GPa the calculations are completely different (compare figure and tables). Below is an example for PUM is from Sun & McDonough (1989).
Request
Is it possible to add the EOS for lower upper mantle phases?
I see that there are some python scripts for doing this, but I haven't looked at them carefully yet. I would be happy to work on this and submit a pull request after a little direction.
Thanks for the consideration!
Buchanan
Example for PUM
Mantle composition
| SiO2 | Al2O3 | CaO | MgO | FeO | K2O | Na2O | TiO2 | O2 | Cr2O3 | H2O | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| PUM | 44.90 | 4.44 | 3.54 | 37.71 | 8.03 | 0.029 | 0.36 | 0.20 | 0.01 | 0.38 | 0 |
Stable solutions
Figure 1: PT diagrams for Primitive Upper Mantle (Sun & McDonough, 1989) showing mineral assemblages predicted by (a) MAGEMin and (b) Perple_X. Greyscale represents the integer indexes of predicted mineral assemblages (see table below).
Density
Figure 2: PT diagrams for Primitive Upper Mantle (Sun & McDonough, 1989) showing density (greyscale in g/cm3) predicted by (a) MAGEMin and (b) Perple_X. Normalised differences and maximum difference gradients are shown in (c, d). Notice the inconsistent phase transitions above 10 GPa.
Stable solutions index
The indexes are generally ordered from low–high pressure conditions. This encoding corresponds to Figure 1 above.
| Index | Perple_X assemblage | MAGEMin assemblage |
|---|---|---|
| 1 | cor cpx o opx ru sp | cpx ol opx ru spn |
| 2 | cpx o opx ru sp | cpx ol opx spn |
| 3 | cpx o opx sp | cpx ol opx pl4T spn |
| 4 | cpx feldspar o opx sp | cpx liq ol opx spn |
| 5 | cpx feldspar melt o opx sp | cpx liq ol opx |
| 6 | cpx melt o opx sp | liq ol opx |
| 7 | cpx melt o opx | liq ol opx spn |
| 8 | melt o opx sp | liq ol spn |
| 9 | melt o opx sp sp | liq ol |
| 10 | melt o sp sp | liq |
| 11 | melt o sp | cpx g ol opx ru spn |
| 12 | melt sp | cpx g ol opx spn |
| 13 | cor cpx gt o opx ru sp | cpx g ol opx ru |
| 14 | cpx gt o opx ru sp | cpx g ol opx |
| 15 | cpx gt o opx sp | cpx g liq ol opx |
| 16 | cpx gt o opx ru | cpx g ol |
| 17 | cpx gt o opx | g liq ol opx |
| 18 | nan | cpx g liq ol |
| 19 | cpx gt melt o opx sp | cpx g ilm ol opx spn |
| 20 | cpx gt melt o opx | liq opx |
| 21 | cpx gt o | cpx g ilm ol opx |
| 22 | cpx gt melt o | cpx_2 cpx_2 g ilm ol opx |
| 23 | gt melt o opx sp | g liq opx |
| 24 | gt melt o opx | cpx_2 cpx_2 g_2 g_2 ilm ol opx |
| 25 | melt opx sp | cpx_2 cpx_2 g_2 g_2 ol opx |
| 26 | gt melt opx sp | cpx_2 cpx_2 g ol opx |
| 27 | gt melt opx | cpx_2 cpx_2 g ol stv |
| 28 | cpx gt hen o opx | cpx g liq ol stv |
| 29 | cpx gt hen o | cpx g liq stv |
| 30 | cpx cpx gt hen o | cpx g ol stv |
| 31 | cpx cpx gt hen o wa | cpx_2 cpx_2 g ol opx stv |
| 32 | cpx gt hen o wa | cpx g ol opx stv |
| 33 | cpx gt iron o opx | cpx g liq opx stv |
| 34 | cpx gt iron melt o opx | cpx g liq ol opx stv |
| 35 | cpx cpx frw gt hen o wa | cpx_2 cpx_2 g ol opx spn stv |
| 36 | cpx cpx frw gt hen hol o | cpx liq ol opx stv |
| 37 | cpx gt hen o opx wa | cpx liq opx stv |
| 38 | cpx gt o opx wa | cpx liq ol opx pl4T spn |
| 39 | cpx cpx gt hen hol o | cpx g liq ol opx_2 opx_2 stv |
| 40 | cpx gt hen hol o | cpx g liq ol opx spn |
| 41 | cpx gt iron o opx wa | cpx g ol opx_2 opx_2 stv |
| 42 | cpx gt hen iron o wa | cpx g ol opx_2 opx_2 |
| 43 | cpx gt hen iron o opx wa | cpx_2 cpx_2 g ol opx_2 opx_2 spn stv |
| 44 | cpx frw gt hen hol o | cpx_2 cpx_2 g ol opx_2 opx_2 stv |
| 45 | cpx gt hen hol iron o | |
| 46 | cpx frw gt hen hol iron o | |
| 47 | cpx cpx frw gt hol mwd stv | |
| 48 | cpx cpx frw gt hen hol mwd | |
| 49 | cpx frw gt hen hol mwd | |
| 50 | cpx gt hen hol iron o wa | |
| 51 | cpx cpx frw gt hol iron mwd stv | |
| 52 | cpx frw gt hol mwd stv | |
| 53 | cpx frw gt hen hol iron mwd | |
| 54 | cpx frw gt hol iron mwd stv | |
| 55 | cpx gt hen hol iron mwd wu | |
| 56 | cpx gt hen hol iron mwd o | |
| 57 | cpx gt hol iron mwd stv wu | |
| 58 | cpx gt gt hen hol iron mwd wu | |
| 59 | cpx gt hen hol iron o opx | |
| 60 | cpx gt hol iron o opx | |
| 61 | cpx cpx gt hol iron mwd stv wu | |
| 62 | cpx gt iron o opx wu | |
| 63 | cpx frw gt hol iron mwd nagt stv | |
| 64 | cpx gt hen hol iron mwd opx wu | |
| 65 | cpx gt hol iron mwd opx wu | |
| 66 | cpx gt hol iron mwd o opx | |
| 67 | cpx gt iron mwd o opx wa | |
| 68 | cpx frw gt hol iron mrw nagt stv | |
| 69 | cpx gt hol iron mwd nagt stv wu | |
| 70 | cpx cpx gt hol iron mwd nagt stv wu | |
| 71 | cpx gt iron mwd opx wa wu | |
| 72 | cpx gt iron o opx wa wu | |
| 73 | cpx gt hol iron mrw nagt stv wu | |
| 74 | cpx gt hol mrw nagt stv wu | |
| 75 | cpx gt iron mwd o opx wa wu | |
| 76 | gt hol iron mwd nagt stv wu | |
| 77 | gt hol iron mrw nagt stv wu | |
| 78 | cpx gt hol iron mwd opx stv wu | |
| 79 | gt hol mrw nagt stv wu | |
| 80 | gt gt hol iron mrw nagt stv wu | |
| 81 | cpx gt hol mrw nagt opx stv wu | |
| 82 | cpv gt hol iron mrw nagt stv wu | |
| 83 | cpv gt hol iron mwd nagt stv wu | |
| 84 | cpv cpx gt iron mwd opx wa wu | |
| 85 | cpv gt iron mwd opx wa wu | |
| 86 | cpx gt gt hol mrw nagt opx stv wu | |
| 87 | cpx gt hol iron maj mwd wu | |
| 88 | cpx gt hol iron maj mwd opx wu | |
| 89 | cpv cpx gt hol iron mwd opx wu | |
| 90 | cpv gt hol iron mwd opx wu | |
| 91 | cpv cpx gt hol iron maj mwd nagt wu | |
| 92 | cpv cpx gt hol iron maj mwd wu | |
| 93 | cpv gt hol iron maj mwd opx wu | |
| 94 | gt hol mrw nagt opx stv wu | |
| 95 | cpv gt hol iron maj mwd nagt wu | |
| 96 | cpv gt hol mrw nagt stv wu | |
| 97 | cpv gt hol mwd nagt stv wu | |
| 98 | cpv gt hol maj mwd nagt wu | |
| 99 | cpv gt hol mak mwd nagt wu | |
| 100 | cpv cpx gt hol mrw nagt opx stv wu | |
| 101 | cpv gt hol mrw nagt opx stv wu | |
| 102 | cpv gt gt hol iron mrw nagt stv wu | |
| 103 | cpv gt hol mak mrw nagt wu | |
| 104 | cpv cpx esk gt hol mrw nagt opx stv wu | |
| 105 | cpv cpx cpx esk hol mrw nagt opx stv wu | |
| 106 | cpv cpx esk hol mrw nagt opx stv wu | |
| 107 | cpv gt gt hol maj mwd nagt wu | |
| 108 | cpv gt hol iron mak mrw nagt wu | |
| 109 | cpv cpx esk hem hol mrw nagt opx stv wu | |
| 110 | cpv cpx esk hem hol nagt opx per stv wu | |
| 111 | cpv gt hol maj mrw nagt wu | |
| 112 | cpv gt gt hol mak mrw nagt wu | |
| 113 | cpv gt hol mpv mrw nagt wu | |
| 114 | cpv gt gt hol mpv mrw nagt wu | |
| 115 | cpv gt hol mpv mwd nagt wu | |
| 116 | cpv gt hol mpv nagt per wu | |
| 117 | cpv gt hol mak mrw nagt opx wu | |
| 118 | cpv gt gt hol mak mrw nagt opx wu | |
| 119 | cpv gt hol mpv mrw nagt opx wu | |
| 120 | cpv cpx esk hem hol nacfb opx per stv wu | |
| 121 | cpv gt hol mpv nagt opx per wu | |
| 122 | cpv gt gt hol mpv mrw nagt opx wu | |
| 123 | cpv gt gt hol mpv nagt per wu | |
| 124 | cpv cpx gt hol nagt opx per stv wu | |
| 125 | cpv cpx esk gt hol nagt opx per stv wu | |
| 126 | cpv cpx gt hol mscf nagt opx stv wu | |
| 127 | cpv cpx gt hol mpv nagt opx per wu | |
| 128 | cpv cpx esk hol nagt opx per stv wu | |
| 129 | cpv cpx esk hol mscf nagt opx stv wu | |
| 130 | cpv cpx esk hol mpv nagt opx per wu | |
| 131 | cpv cpx esk gt hol mpv nagt opx per wu | |
| 132 | cpv cpx esk hol mpv mscf nagt opx wu | |
| 133 | cpv gt hol mpv nacfb opx per wu | |
| 134 | cpv cpx cpx esk hol nagt opx per stv wu | |
| 135 | cpv cpx esk hol nacfb opx per stv wu | |
| 136 | cpv cpx gt hol mpv nacfb opx per wu | |
| 137 | cpv cpx esk hol mpv nacfb opx per wu | |
| 138 | cpv cpx esk hol mscf nacfb opx stv wu | |
| 139 | cpv cpx geik gt hol mpv nacfb opx per wu | |
| 140 | cpv geik gt hol mpv nacfb opx per wu | |
| 141 | cpv cpx esk hem hol mscf nacfb opx stv wu | |
| 142 | cpv cpx esk hol mpv mscf nacfb opx wu | |
| 143 | cpv cpx esk geik hol mpv nacfb opx per wu | |
| 144 | cpv cpx esk hem hol mpv mscf nacfb opx wu | |
| 145 | cpv cpx esk hem hol mpv nacfb opx per wu | |
| 146 | cpv esk geik gt hol mpv nacfb opx per wu | |
| 147 | cpv esk geik hem hol nacfb opx per stv wu | |
| 148 | cpv esk geik hem hol mpv nacfb opx per wu | |
| 149 | cpv esk geik hem hol mpv mscf nacfb opx wu | |
| 150 | cpv esk geik hem hol mscf nacfb opx stv wu |
Perplex build
I use hp634ver.dat with the following input.txt to build the Perple_X model with ./build < input.txt. The perplex_option.dat is only modified to set x_nodes and y_nodes to 32 128:
PUM
hp634ver.dat
perplex_option.dat
N
2
N
N
N
SiO2
Al2O3
CaO
MgO
FeO
K2O
Na2O
TiO2
O2
Cr2O3
H2O
5
N
2
773 2273
10000 280000
Y
44.90 4.44 3.54 37.71 8.03 0.029 0.36 0.20 0.01 0.38 0 82.4
N
Y
N
enL
fo8L
foTHL
faTL
fa8L
fbi
perL
anL
ctjL
qjL
hmL
hmWL
hmTHL
tiL
tiWL
tiTHL
hemL
ruL
Y
solution_model.dat
feldspar
Cpx(HGP)
melt(HGPH)
O(HGP)
Sp(HGP)
Gt(HGP)
Opx(HGP)
PUM
Hi Buchanan,
The igneous database (Holland et al., 2018) has been calibrated roughly from 700 to 2000°C and 0 to 5-6 GPa and do not take into account phase change deeper than this.
One of the main difference between Perple_X and MAGEMin is that we only provide the database as published (including corrective updates) and do not offer the possibility to add phases that were not included for the inversion, at least natively.
However, we are working on adding the Stixrude database that would work for these deeper conditions. It will take some time before the database is made public, though.
That being said, if you want to consider the deep mantle pure phases present in the ts34 database and hope for the best, this can be done on your branch/local version of the code. Note that the igneous database has not been inverted with these phases and the results may likely be wrong.
To add the pure phases for deeper conditions, you will have to change line 309 of the initialize.h file, add the pure phases you want to the list, then modify the counter of pure phases line 306 and 280. Then recompile. Again we cannot guarantee that the results will make sense.
Hope this helps,
Hi Nicolas,
Thanks for the clarification and suggestions.
For now I am mostly interested in achieving consistency between MGM and PPX, with less focus on accuracy of the predicted mineral assemblages, so I will try your suggestion and see how far it goes.
I will look forward to the Stixrude database and more future releases!
Best,
Buchanan