SiO2 content of liquids rhyolite-MELTS (GUI) versus pHMELTS (alphaMELTS)

[Ri Cao]

Dear Paula,
This is Ri. Sorry for my interruption again and hope you are doing well. Thanks for your advice regarding MELTS running. I have a naive question there regarding the MELTS download:

Recently, I just have comparisons between Rhyolite-MELTS and pHMELTS regarding SiO2 content and solidus temperature of liquids. I was wondering are there still any official websites/places for the download of pHMELTS software? 

After reading the documentation, just to double-check does that mean I could run this pHMELTS based on alphamelts as well? Thank you very much and I look forward to your response.

Sincerely,
Ri Cao

[Paula]

Yes, alphaMELTS is the pHMELTS software. At the time of the pHMELTS paper it was called Adiabat_1ph, but it was later renamed:

https://magmasource.caltech.edu/forum/index.php/topic,123.msg261.html#msg261

At the moment alphaMELTS 2 doesn't have the pHMELTS model. Much of the necessary coding is done and hidden, e.g. to buffer aH2O, but I've not finished implementing trace elements (strictly trace species). You can run the pHMELTS model in alphaMELTS 1.9, though.

Best wishes,
Paula

[Ri Cao]

Dear Paula,
Thank you very much for your prompt response and useful clarification.

One thing I was wondering is that since I just found a large difference for the solidus temperature and SiO2 content that estimated by Rhyolite-MELTS 1.0.2 version and pHMELTS for the following two compositions at higher pressures (i.e., 1 GPa):
Composition 1: SiO2: 46.64wt.%  TiO2 1.65 wt.%  Al2O3: 16.34 wt.%  FeO: 9.62 wt.%  MnO: 0.21 wt.%  MgO: 11.79 wt.%  CaO: 7.34 wt.% Na2O: 2.07 wt.% K2O: 4.14 wt.% H2O:0.2 wt.%

Composition 2 (close to MORB): SiO2: 49.69 wt.% TiO2: 1.28 wt.% Al2O3: 18.26 wt.% MnO: 0.16 wt.%  MgO: 8.26 wt.%  CaO: 10.51 wt.% Na2O: 2.45 wt.% K2O: 0.2 wt.% H2O: 0.2 wt.%.


I was wondering do you have some recommendations on which versions of MELTS should be used for compositions 1, 2, separately? (i.e., Composition 1: pHMELTS, composition 2: rhyolite-MELTS)? Any thoughts on this will be very helpful! Thank you very much and I look forward to your response.

Sincerely,
Ri Cao

[Paula]

Which underlying thermodynamic model are you using in pHMELTS? MELTS or pMELTS? MnO isn't calibrated in the pMELTS liquid, so you should remove that for comparison if you are using pMELTS. Do you have ALPHAMELTS_DO_TRACE_H2O on?

Also at higher pressures it is better to treat potassium as a trace element, as there are few solid phases for K2O to go into in the MELTS system.

[Ri Cao]

Dear Paula,
Thank you very much for your response.

Since I am a completely fresher user in alphamelts, pHMELTS, I just got the alphamelts_default_env.txt file there:

%%%%%
!Default values of environment variables (version 1.6)
!Variables preceeded by '!' are 'unset' (i.e. 'false')

ALPHAMELTS_VERSION pHMELTS
!ALPHAMELTS_OLD_GARNET true
ALPHAMELTS_OLD_SPINEL true
!ALPHAMELTS_OLD_BIOTITE true
!ALPHAMELTS_2_AMPH true
!ALPHAMELTS_NO_CHLORITE true
ALPHAMELTS_MODE isobaric
!ALPHAMELTS_PTPATH_FILE filename
ALPHAMELTS_DELTAP 0
ALPHAMELTS_DELTAT -5
ALPHAMELTS_MAXP +10000
ALPHAMELTS_MINP +0
ALPHAMELTS_MAXT +1400
ALPHAMELTS_MINT +600
!ALPHAMELTS_ALTERNATIVE_FO2 true
!ALPHAMELTS_LIQUID_FO2 true
!ALPHAMELTS_IMPOSE_FO2 true
!ALPHAMELTS_FO2_PRESSURE_TERM true
!ALPHAMELTS_CONTINUOUS_MELTING true
ALPHAMELTS_MINF 0.005
!ALPHAMELTS_MINPHI 0.002
!ALPHAMELTS_CONTINUOUS_RATIO float
!ALPHAMELTS_CONTINUOUS_VOLUME true
!ALPHAMELTS_FRACTIONATE_SOLIDS true
ALPHAMELTS_MASSIN  0.001
ALPHAMELTS_FRACTIONATE_WATER true
!ALPHAMELTS_MINW float
!ALPHAMELTS_FRACTIONATE_TARGET true
!ALPHAMELTS_MGO_TARGET 8.0
!ALPHAMELTS_MGNUMBER_TARGET float
!ALPHAMELTS_ASSIMILATE true
!ALPHAMELTS_FLUX_MELTING true
!ALPHAMELTS_DRY_ITER_PATIENCE 100
!ALPHAMELTS_DO_TRACE true
!ALPHAMELTS_DO_TRACE_H2O true
!ALPHAMELTS_HK_OL_TRACE_H2O true
!ALPHAMELTS_HK_PXGT_TRACE_H2O true
!ALPHAMELTS_2X_OPX_TRACE_H2O true
!ALPHAMELTS_TRACE_DEFAULT_DPTX true
!ALPHAMELTS_TRACE_NORMALIZATION integer
!ALPHAMELTS_TRACE_INPUT_FILE filename
!ALPHAMELTS_TRACE_USELIQFEMG true
!ALPHAMELTS_MULTIPLE_LIQUIDS true
!ALPHAMELTS_FRACTIONATE_SECOND_LIQUID true
!ALPHAMELTS_FOCUS true
!ALPHAMELTS_FOCUS_FACTOR float
!ALPHAMELTS_ADIABAT_FILE true
!ALPHAMELTS_CELSIUS_OUTPUT true
!ALPHAMELTS_SAVE_ALL true
!ALPHAMELTS_SKIP_FAILURE true
!ALPHAMELTS_FAILED_ITER_PATIENCE 10
!ALPHAMELTS_INTEGRATE_FILE filename
!ALPHAMELTS_LATENT_HEAT true
!ALPHAMELTS_QUICK_OUTPUT true

! Use Celsius instead of Kelvin
ALPHAMELTS_CELSIUS_OUTPUT true
%%%%

Regarding potassium at higher pressures, does that mean I should change the potassium from ''Initial Composition: K2O 0.2000'' to something like '' Initial Trace: K ''?

I am trying to reproduce the data from this paper (https://agupubs.onlinelibrary.wiley.com/doi/full/10.1002/jgre.20094), where all the calculation for batch melting and fractional crystallisation for Venera 13, 14 are based on pHMELTS. For fractional crystallisation, my current setting in pHEMTLS could almost reproduce their data well. However, I am a little bit confused there regarding batch melting calculation since I cannot reproduce the chemical composition of liquid appropriately. Any thoughts on this will be very helpful! Thank you very much.

Sincerely,
Ri Cao

[Paula]

Hi Ri,

It's not completely clear what he did. pHMELTS isn't the program, it is a thermodynamic model based on MELTS / pMELTS (see Asimow et al 2004). That's why I asked which underlying model you were using. The paper doesn't say whether MELTS or pMELTS was used, But as the pMELTS paper isn't cited, and they are doing calculations at <= 1 GPa, I assume it is MELTS.

[Also, the quoted list of oxides that the model is calibrated for is for MELTS, but people often mistakenly put MnO and P2O5 in pMELTS calculations so that's a less reliable indicator of which model was used.]

alphaMELTS (originally Adiabat_1ph) is the name for the software that is a front end to MELTS (in alphaMELTS 1.X, updated to rhyolite-MELTS in alphaMELTS 2), pMELTS and pHMELTS.

The default thermodynamic model for alphamelts is pMELTS. So, as alphaMELTS 1.X won't recognize "pHMELTS" as an option for ALPHAMELTS_VERSION, you are running pMELTS.

You run the pHMELTS model by choosing an underlying thermodynamic model - usually pMELTS but it doesn't have to be (there's some discussion of this in the documentation) - and then setting ALPHAMELTS_DO_TRACE_H2O to true.

So I would try replacing "pHMELTS" with "MELTS" and see if the results match better. Also, yes, you are correct about how to treat potassium as a trace element, but it looks like it was treated as an oxide in the calculations in the paper.

Paula

[Ri Cao]

Dear Paula,
Thank you very much for your response and clarification. Sorry for my slow response and everything is running slowly here due to COVID.

Yes, I tried to replace ''PHMELTS'' with ''MELTS'' and everything works now. I also got a reply from the author of that paper and yes, he suggested the same thing ( he used a DOS-based ‘normal’ MELTS (Smith and Asimow, 2005, G3, Q02004, doi:10.1029/2004GC000816). Thank you very much.

Sincerely,
Ri Cao

[Ri Cao]

Dear Paula,
Sorry for my interruption again and hope you are doing well.

I just realized that there is another MELTS workshop for the IAVCEI conference in New Zealand, early next year. I was wondering how shall I sign up for this workshop? Are there any links for registration and how much will that cost? Thank you very much and I look forward to your response.

Sincerely,
Ri Cao

[Paula]

Hi Ri,

I'm not sure. They haven't updated the website fully. I guess you can add the workshop when registering for the conference, but I haven't tried yet. To be honest, I'm still tying up loose ends from the recent workshop at Goldschmidt!

The full price is supposed to be 160 NZD. Students and early-career researchers (as defined here) should get a reduced rate of 60 NZD. I will follow up next week to see how you get that rate.

Best wishes,
Paula

[Ri Cao]

Dear Paula,
Great, thanks for the clarification.

I also got some replies from the conference organizer as well. Sorry for missing the Goldsmith workshop since I was ill at that time period.

Hopefully and definitely, I will be able to attend the next MELTS conference in Newzeland. Please keep me in the loop if there is anything new that comes through. Thank you very much.

Best regards,
Ri Cao