Include MORE simulation results for Earth model

[lena-noack]

No description provided.

[lena-noack]

Commit of first TRAPPIST-1 example models; also pull request of Earth models was not yet added several months ago?

[nichollsh]

Hi Lena, thank you for submitting these data! It's great to have more models to compare with. I am happy to give a first pass at reviewing this PR but would also feel more comfortable if one of the @projectcuisines/chili_admin could also take a look at a later stage.

However, there are a number of issues that need to be addressed here. I am confused about what the model is doing. Some of the data values seem incorrect but it's hard for me to know what to expect from MORE. Perhaps you can clarify these points below?

@lena-noack @philippbaumeister @timlichtenberg

(1) Static vs evolutionary

It was my understanding that MORE is a static model, not an evolutionary model, since it has no notion of time evolution. You mention this in the notes too. If this is the case, why are the data submitted in the format of the evolution codes, rather than in the protocol-defined format for static models? That is: the protocol says that static models should calculate the state at specific times derived from the evolution models (via the Exoplanet paper simulations).

Could you explain why the data are formatted in the evolution-model format rather than the static-model format?

You can see how to do this here: https://github.com/projectcuisines/chili/blob/main/intercomparison/README.md#static-models

(2) Missing values

• The output CSV files are missing a lot of values and some of the values do not make sense to me. For example, the T_surf(K) doesn't change between the rows, even though the melt fraction is changing. Seems physically inconsistent?

• If this is being interpreted as an evolution calculation, why does the melt fraction increase down the rows?

• The fO2_solid(bar) and fO2_melt(bar) values are all constant and seem to be in the wrong units. They are fixed at a value of 4 (I guess meaning IW+4) but the protocol says that these should be absolute $fO_2$ in units of bar.

• The Venus outputs and several of the exoplanet cases are missing.

• The file names are mis-formatted. For example, there is a file called ‎evolution-more-evolution-more-trappist1e-data-data.csv which I think should be called evolution-more-trappist1e-data.csv. Although, again, this is related to my point above about static/evolutionary.

(3) Missing profiles?

The CHILI protocol specifies that static models are to ideally provide atmosphere profiles (see here). I do not see these in your output data. If your model doesn't calculate these, then that's fine, but the output format should still follow the static-model specification.

(4) Model description

The model description the -notes.txt file is quite short. The other CHILI models usually provide a citation in this case, or point to their online documentation: e.g. LINCS has some references and MOAI has a website.

Could you perhaps expand your notes file and/or add some references so it's clearer to understand what the model is doing? Maybe this will clarify my points of confusion above.

[lena-noack]

Hi Harrison, Thanks a lot for the detailed check. Here are my replies: 1) Static vs evolutionary This is exactly the reason for the misunderstanding that we had about the question when to submit our models. MORE is an evolutionary model, but instead of calculating the change of surface temperature and expected cooling of the interior from that (which depends on several factors), it instead ASSUMES that the magma ocean freezes and based on an enforced freezing from iteration step to iteration step, it then calculates the redistribution of volatiles between the interior and surface. Therefore there is no actual time dependence, as the time does not play a role, only the degree of freezing matters. This is why I would call it a "time-independent evolution model", it is definitely not what is meant with a static model here. Comparing the volatiles redistribution between MORE and other codes when comparing not against time, but against melt fraction, leads to quite comparable results, showing that the quite different approach of the MORE model still works (unless atmospheric erosion becomes strong). So it is an evolutionary model, but in a different way than what was meant in CHILI to describe time-dependent models, and this is why we were first asked to not submit our results, which then later turned out to be exactly this misunderstanding, that our model is also not a static model (no atmosphere profile is calculated, and surface temperature is indeed set as a boundary condition). The different philosophy with which the MORE model was developed was to allow for more easy parameter studies (different planetary sizes, redox states, compositions, more gases than studied here etc.) independent of the star, whereas the other CHILI models instead focus on the star-planet connection but then also always have to assume stellar properties. But this also means that the MORE model is not suitable for all exoplanets, especially not for close-in exoplanets. 2) Missing values - T_surf(K) indeed is constant as it is not calculated (and actually not really used, only to have a temperature as input for atmospheric chemical equilibrium), the cooling of the magma ocean is enforced via a decreasing melt fraction - melt fraction: The rows are apparently just flipped up-down, so the simulation starts at the bottom (melt fraction 1) and stops at the top (melt fraction stops when falling below 5%). - fO2_solid(bar) and fO2_melt(bar): indeed these are constant at IW+4. I thought I saw other codes also give the IW variation, also as the fO2 value depends a lot on local pressure/temperature and nees to be averaged, so a buffer value seemed more logical to me, but I can change that of course to averaged $fO_2$ in units of bar. Or it seems those not changing redox during crystallization just take NaN. I have a model variation with variable redox state but it is still in the testing/benchmarking phase and not ready yet to be added here. - The Venus models go again more into the direction of atmosphere losses/different atmospheric conditions, so here MORE is not really valid anymore, and it is more appropriate to only apply it to Earth. The only change in my code from Earth to Venus is the size and mass of the planet, which seemed a bit too boring to add here... But of course I can upload also a Venus case, assuming also here some weak H loss similar to the Trappist-1b case. - Exoplanet cases: So these are first just some test cases to see if the simulations are at all comparable to what the time-dependent models give, as I do not have a time-dependent atmosphere loss included (I can only scale it a bit up or down, as described in the notes). But looking at a couple of other folders in the CHILI GitHub, it seems that others also just added one or two simulations? So which ones are missing? The alpha does not make sense for MORE to calculate, 1b is already highly fictional in terms of atmospheric loss... Again as I do not calculate atmospheric temperatures, normally only 1e would make sense in the comparison, but this is essentially the Earth-case from the solar system study - File names: indeed I had a copy and past error when chaning the output name and added "evolution-more...data" twice by accident. Also the solar system ones still have "model1" in the name from some time ago, where I saw this naming also in one of the other folders and copied it wrong. This can of course easily be changed. 3) Missing profiles? As MORE is an evolutionary model (just not strictly time-dependent) I can not provide the static model files. (4) Model description For the notes.txt file in the output, I also compared first to what other codes were writing and it seems to not bee much more information. The main parameter description is as described in the protocol in the inputs/more folder, though I just realized that I forgot to add there the Trappist files, will add them asap. For a more detailed description, there is a paper in preparation which I mentioned earlier, which could be a half-CHILI paper with opt-in for those interested, where the model (especially since the philosophy is different to other magma ocean models) would be described in more detail. I can of course copy the model description in bullet points here already. So I will - update the file names, add Trappist input files - flip the data up-down in the files - check the fO2 calculation, maybe just add NAN as it stay constant in my model - add more description to the notes file I will push the changes then soon, I hope it does not create problems with the earlier CHILI pull request... Best, Lena

…

-- Prof. Dr. Lena Noack Freie Universität Berlin Department of Earth Sciences Malteserstr. 74-100 Room D211 - Building D 12249 Berlin Tel.: +49 (0) 30 838 636 94 Fax: +49 (0) 30 838 4 636 94 E-mail: ***@***.***

________________________________ Von: Harrison Nicholls ***@***.***> Gesendet: Dienstag, 16. Juni 2026 13:32 An: projectcuisines/chili Cc: Noack, Lena; Mention Betreff: Re: [projectcuisines/chili] Include MORE simulation results for Earth model (PR #76) @nichollsh requested changes on this pull request. Hi Lena, thank you for submitting these data! It's great to have more models to compare with. I am happy to give a first pass at reviewing this PR but would also feel more comfortable if one of the @projectcuisines/chili_admin<https://github.com/orgs/projectcuisines/teams/chili_admin> could also take a look at a later stage. However, there are a number of issues that need to be addressed here. I am confused about what the model is doing. Some of the data values seem incorrect but it's hard for me to know what to expect from MORE. Perhaps you can clarify these points below? @lena-noack<https://github.com/lena-noack> @philippbaumeister<https://github.com/philippbaumeister> @timlichtenberg<https://github.com/timlichtenberg> (1) Static vs evolutionary It was my understanding that MORE is a static model, not an evolutionary model, since it has no notion of time evolution. You mention this in the notes too. If this is the case, why are the data submitted in the format of the evolution codes, rather than in the protocol-defined format for static models? That is: the protocol says that static models should calculate the state at specific times derived from the evolution models (via the Exoplanet paper simulations). Could you explain why the data are formatted in the evolution-model format rather than the static-model format? You can see how to do this here: https://github.com/projectcuisines/chili/blob/main/intercomparison/README.md#static-models (2) Missing values * The output CSV files are missing a lot of values and some of the values do not make sense to me. For example, the T_surf(K) doesn't change between the rows, even though the melt fraction is changing. Seems physically inconsistent? * If this is being interpreted as an evolution calculation, why does the melt fraction increase down the rows? * The fO2_solid(bar) and fO2_melt(bar) values are all constant and seem to be in the wrong units. They are fixed at a value of 4 (I guess meaning IW+4) but the protocol says that these should be absolute $fO_2$ in units of bar. * The Venus outputs and several of the exoplanet cases are missing. * The file names are mis-formatted. For example, there is a file called ‎evolution-more-evolution-more-trappist1e-data-data.csv<https://github.com/lena-noack/chili/blob/b54432a6249c0332ebccd829751bff2d82b2ffd4/intercomparison/outputs/more/evolution-more-evolution-more-trappist1e-data-data.csv> which I think should be called evolution-more-trappist1e-data.csv. Although, again, this is related to my point above about static/evolutionary. (3) Missing profiles? The CHILI protocol specifies that static models are to ideally provide atmosphere profiles (see here<https://github.com/projectcuisines/chili/blob/main/intercomparison/README.md#static-models>). I do not see these in your output data. If your model doesn't calculate these, then that's fine, but the output format should still follow the static-model specification. (4) Model description The model description the -notes.txt file is quite short. The other CHILI models usually provide a citation in this case, or point to their online documentation: e.g. LINCS<https://raw.githubusercontent.com/projectcuisines/chili/refs/heads/main/intercomparison/outputs/lincs/evolution-lincs-notes.txt> has some references and MOAI<https://planetomoai.readthedocs.io/en/latest/foreword/MOAI.html> has a website. Could you perhaps expand your notes file and/or add some references so it's clearer to understand what the model is doing? Maybe this will clarify my points of confusion above. — Reply to this email directly, view it on GitHub<#76?email_source=notifications&email_token=BTP62FVHGPNQAJRSHPYVITL5AEV6VA5CNFSNUABKM5UWIORPF5TWS5BNNB2WEL2QOVWGYUTFOF2WK43UKJSXM2LFO4XTINJQGU4TCOBZHE4KM4TFMFZW63VHNVSW45DJN5XKKZLWMVXHJLDGN5XXIZLSL5RWY2LDNM#pullrequestreview-4505918998>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/BTP62FQWPBIZ5U475O667AD5AEV6VAVCNFSNUABFKJSXA33TNF2G64TZHM4TSMJYHE4TSMZSHNEXG43VMU5TIMZZGI3DONJQGQ4KC5QC>. Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS<https://github.com/notifications/mobile/ios/BTP62FS7M3PXMZCFIJGXAVL5AEV6VA5CNFSNUABKM5UWIORPF5TWS5BNNB2WEL2QOVWGYUTFOF2WK43UKJSXM2LFO4XTINJQGU4TCOBZHE4KM4TFMFZW63VHNVSW45DJN5XKKZLWMVXHJKTGN5XXIZLSL5UW64Y> and Android<https://github.com/notifications/mobile/android/BTP62FRUPGOFVACVWIBLUSD5AEV6VA5CNFSNUABKM5UWIORPF5TWS5BNNB2WEL2QOVWGYUTFOF2WK43UKJSXM2LFO4XTINJQGU4TCOBZHE4KM4TFMFZW63VHNVSW45DJN5XKKZLWMVXHJLTGN5XXIZLSL5QW4ZDSN5UWI>. Download it today! You are receiving this because you were mentioned.Message ID: ***@***.***>