As the planetary science community sets its sights on the Moon, the existence of an open-source, up-to-date, and user-friendly modeling tool for lunar rocks is critical to maximizing the scientific return of ongoing and upcoming lunar missions (e.g., Artemis, PRISM, CLPS). While the creation of new code from the ground up is an important aspect of modern computational petrology, here we advocate for the modernization of legacy code, the results of which dominate the literature and shape our current understanding of geologic processes across multiple scales. Modernization of legacy code is critical as it enables the community to put new model results in the context of modern consensus gentium. Here we discuss how modern best practices for code development, publishing, and maintenance should be applied to upgrading legacy code, using lunar petrologic models as an example. We highlight critical gaps in our ability to model lunar processes that could be filled simply with updated modeling tools (i.e., where underlying experimental and analytical data already exist but are not incorporated into existing modeling tools).

Current modeling tools developed specifically for lunar compositions are sparse and can contain outdated parameterizations. One critical knowledge gap is our ability to model silicic lunar magmas, which are evidenced in nature by felsic fragments in returned Apollo samples and silica-rich volcanic domes identified on the borders of lunar mare by remote sensing. The most popular tool for modeling lunar magmas is MAGPOX, born from a series of FORTRAN scripts and ported to MATLAB, which is underpinned by an exclusively basaltic database. MAGPOX requires the crystallization of olivine on the liquidus, and thus has limited application to the full compositional diversity of lunar magmas. rhyolite-MELTS has been used to model silicic magmatism on the Moon, but its use typically requires additional experimental work given that the MELTS database is biased towards terrestrial rocks with lower FeO, higher alkalis, and higher fO₂ than lunar rocks. Notably, the MELTS database includes the same published basaltic lunar rocks used in MAGPOX regressions and so should be trustworthy for silica-poor lunar magmas. Still, the adoption of MELTS by the lunar community requires extensive testing against MAGPOX, PERPLE_X, and large lunar experimental databases.

Before we can even begin to update the parameterizations underpinning MAGPOX or efficiently compare MAGPOX to other models, a modern code library is required to perform adequate testing and benchmarking. In this talk we will explore the state of lunar petrologic modeling, what can be done now, and how we can best bring it into the 21^st century.

How to cite: Iacovino, K. and Prissel, K.: Modeling lunar magmas in the Artemis Era, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-2383, https://doi.org/10.5194/egusphere-egu23-2383, 2023.

Despite the outstanding progress in computer technology and experimental facilities, understanding melting processes and solid-melt phase equilibria at planetary conditions is still an open challenge. In this work a modern computational approach to predict melting phase relations at HP-HT by a combination of first principles DFT and MD calculations, polymer chemistry and equilibrium thermodynamics is presented and discussed. The adopted theoretical framework is physically-consistent and allows to compute multi-component phase diagrams relevant to planetary interiors in a broad range of P-T conditions by a convex-hull algorithm based on the simplex method for Gibbs free energy minimisation. The calculated phase diagrams are in turn used as a source of information to gain new insights on both present-day and early Earth melting processes. Some examples of application of the above method to the CaO-MgO-Al₂O₃-SiO₂ system (CMAS) and relevant ternary and binary subsystems highlights as pressure effects are not only able to change the nature of melting of some minerals (like olivine and pyroxene) from eutectic to peritectic (and vice versa), but also simplify melting relations by drastically reducing the number of phases with a primary phase field at HP-HT conditions. Since the volume-pressure integral contribution to Gibbs free energy become relevant at planetary interior conditions, special attention must be paid to the choice of the P-V-T EoS formalism in order to avoid physical unsoundness or spurious effects in thermodynamic properties (e.g. negative thermal expansion). Ab initio-assisted computational thermodynamics is thus outlined as the main route towards the future development of physically-consistent (besides internally-consistent) thermodynamic databases for global-scale planetary investigations.

Financial support by the Italian Ministry of University and Research (MIUR PRIN 2017, Project 2017KY5ZX8 and MIUR PRIN 2020, Project 202037YPCZ) is warmly acknowledged.   

How to cite: Belmonte, D., La Fortezza, M., and Menescardi, F.: Ab initio-assisted computational thermodynamics: a modern approach to phase diagram calculation at planetary conditions, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-14346, https://doi.org/10.5194/egusphere-egu23-14346, 2023.

In contrast to thermodynamic models for metamorphic mineral solutions, which often concern quartz-saturated assemblages, the phase equilibria of feldspars, feldspathoids and melilite in the K₂O-Na₂O-CaO-Al₂O₃-SiO₂-H₂O space are much less established. They provide basis for classification and interpretation of alkaline silica-undersaturated rocks and essential constraints for developing normative calculation and classification algorithms. The existing normative schemes (CIPW, Müller, Le Bas, Rittmann and Curie norms) do not provide satisfactory treatment of nepheline, leucite, clinopyroxene and melilite stabilities and equilibria. In many conventional schemes (1) anorthite persists to very low silica activities despite of its instability and incompatibility with melilite; (2) normative larnite is a principal indicator of melilite presence but provides a poor chemical proxy; (3) compatibilities between alkali feldspar, leucite and nepheline solid solutions are incorrectly predicted (e.g., leucite-albite, feldspar-melilite); (4) desilication steps do not uniquely follow decreasing activity of silica. As a consequence, the norms applied to alkaline silica-undersaturated igneous rocks do not provide unique or correct view of mineral assemblages and their chemographic relations. In this contribution we explore phase equilibria and compatibility relations in the system SiO₂-CaAl₂O₄-NaAlSiO₄-KAlSiO₄-H₂O. We have adopted the thermodynamic models for nepheline, leucite and kalsilite solutions (Sack & Ghiorso 1998). The order-disorder models have been converted to the relevant sets of linearly independent end-members including ordered or anti-ordered intermediate members with macroscopic Margules parameters. These models were designed as transferable between Berman, Holland-Powell or other end-member datasets. In the composition space SiO₂-NaAlSiO₄-KAlSiO₄ at P = 1 bar and T = 1000 ^oC, alkali feldspar is compatible with nepheline or leucite, separated by an invariant composition Or₅₁Ab₄₉. Consequently, sodic rocks, with molar Na/(Na+K) > 0.75, contain the anorthoclase + nepheline assemblage, whereas with increasingly potassic character, the rocks contain leucite + nepheline, leucite + tetrakalsilite or leucite + kalsilite; the SiO₂-richer assemblages consist of leucite + sanidine. With decreasing temperature, tetrakalsilite becomes unstable and is replaced by K-bearing nepheline + kalsilite. Furthermore, the invariant feldspar composition separating the nepheline vs. leucite assemblages rapidly migrates towards sanidine (Or₉₀Ab₁₀ at 800 ^oC and 1 bar). As a consequence, decreasing crystallization temperature favors the assemblage of alkali feldspar and nepheline (over leucite) for a wide range of bulk Na/(Na+K) ratios. This explains the rarity of sodic feldspar + leucite assemblages in nature. At feldspar subsolvus temperatures, both albite and orthoclase coexist with nepheline, and leucite becomes Na-poor and eventually breaks down to kalsilite and orthoclase. This is consistent with frequent replacement in nature of leucite by secondary products, in particular analcime. The compatibility relations in the SiO₂-NaAlSiO₄-KAlSiO₄ system are conveniently delineated by several intermediate members, e.g., Or₇₅Ab₂₅ or Ne₆₀Ks₄₀, which are used in a new algorithm for normative classification and interpretation of alkaline undersaturated rocks.

References

Sack R.O., Ghiorso M.S., 1998. Contrib. Mineral. Petrol. 130, 256-274.

How to cite: Dolejš, D.: Magmatic equilibria between feldspathoids and feldspars: Implementation of thermodynamic models and implications for norm calculation and petrographic interpretation, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-9700, https://doi.org/10.5194/egusphere-egu23-9700, 2023.

Fluid release from hydrated oceanic lithosphere in subduction zones is a key process in the deep earth water cycle. During prograde metamorphism a channelized vein network forms in the dehydrating rock that allows efficient fluid release from the slab into the mantle wedge. The formation of such a vein network is a multiscale process that occurs over a wide range of time and length scales. Previous studies as well as field observations of exhumed meta-serpentinites suggest that the processes governing rock dehydration shift from chemical to mechanical processes going from small to large scales.

To investigate the behavior of a dehydrating slab over this wide range of scales we present a multiscale dataset that includes field-based observations from m to sub-µm scale of a representative serpentinite from the Mirdita ophiolite (Albania). This ophiolite has experienced seafloor alteration, but has not been metamorphosed at conditions that would cause any dehydration. We use these data as input for thermodynamic equilibrium calculations to investigate the effect of chemical heterogeneities in the bulk rock composition while the PT-conditions will be increased following a typical subduction zone geothermal gradient.

For upscaling we perform the calculations at various effective thermodynamic domain sizes (20-100 µm), showing that anisotropic chemical heterogeneities lead to heterogeneous porosity formation on all scales. Bloch et al. (2018) found that for vein-like porosity structures the percolation threshold of an effective bulk media may be reached at a porosity as low as 10^-3. Therefore, the anisotropic porosity structure formed by chemical heterogeneities leads to a high connectivity even at low porosities and thus allows efficient fluid flow. Accordingly, going to even larger scales we can use these findings to describe the lithologies found in the field as effective bulk media with an effective fluid flow. This allows us to investigate fluid release from the dehydrating slab on the km-scale by reactive porosity waves using a numerical model.

Bloch, W., John, T., Kummerow, J., Salazar, P., Krüger, O. S., & Shapiro, S. A. (2018). Watching Dehydration: Seismic Indication for Transient Fluid Pathways in the Oceanic Mantle of the Subducting Nazca Slab. Geochemistry, Geophysics, Geosystems, 19(9), 3189–3207. https://doi.org/10.1029/2018GC007703

How to cite: Huber, K., Vrijmoed, J. C., and John, T.: Serpentinite dehydration as a multiscale process, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-14500, https://doi.org/10.5194/egusphere-egu23-14500, 2023.

Explorations of mineral compositions aiming to reveal complex magmatic processes in melts have proliferated with the growing accessibility of geochemical datasets through databases including PetDB, LEPR/TraceDs, and GEOROC and of computational methods. The generation and continuous quality assurance of mineral data in these databases requires significant human intervention and individual post-processing. One major problem is that minerals may be misclassified (i.e., a compiled dataset of clinopyroxenes may contain some amphiboles), and compilations may contain poor-quality electron microprobe (EPMA) analyses (with low totals, low cation sums, or poor correspondence to the theoretical stoichiometry of a mineral phase). At the moment, individual studies compiling geochemical datasets for specific tectonic settings [1] or calibrating thermobarometers based on mineral-melt equilibrium [2] tend to apply their own filters. With a push for a more consistent approach, we create a new open-source Python package called MIN-ML (MINeral classification using Machine Learning) for classifying common igneous minerals based on oxide data collected by EPMA, with functions for calculating stoichiometries and crystallographic sites based on this classification. Utilizing this package allows for the identification of misclassified mineral phases and poor-quality data. We streamline data processing and cleaning to allow for the rapid transition to usable data, improving the utility of data curated in these databases and furthering computing and modeling capabilities. 

While mineral identification and classification are obviously critical to the success of computational methodologies and machine learning (ML) applied to these large datasets, the question of how to best classify minerals from EPMA analyses comes to the fore. We approach this question by exploring and developing ML workflows, both supervised (classification algorithms) and unsupervised (dimensionality reduction and clustering). Unsupervised methods including autoencoders, a type of artificial neural network, present the opportunity to classify minerals with little a priori information. Autoencoders pair two neural networks with an encoder, compressing input data to a dimensionality-reduced latent representation, and a decoder, expanding latent representations to reconstruct the input and minimize loss. We present a novel autoencoder model aimed at meaningfully representing EPMA analyses of minerals in latent space, investigating the relationships between mineral phases, and performing classifications of these minerals. The model is trained with newly compiled datasets of twelve igneous mineral phases on thousands to tens of thousands of analyses per phase – across tectonic settings to train these ML models. The autoencoder is applied to datasets of mineral analyses from PetDB, LEPR, and GEOROC to evaluate model performance and show significant improvements in mineral phase segregation and classification, critical to rigorous dataset quality control and future integration into data processing routines. 

[1] Gale, A., et al., The mean composition of ocean ridge basalts. Geochemistry, Geophysics, Geosystems 14, 489-518 (2013).

[2] Petrelli, M., et al., Machine learning thermobarometry: Application to clinopyroxene-bearing magmas. JGR: Solid Earth 125, e2020JB020130 (2020).

[3] Lehnert, K. A., et al., 2022, IEDA2: Evolving EarthChem, LEPR/TraceDs, and SESAR into a Next Generation Data Infrastructure for Data-Driven Research Paradigms in Geochemistry, Petrology, and Volcanology, in 2022 Goldschmidt Conference.

How to cite: Shi, S., Wieser, P., Lehnert, K., and Profeta, L.: MIN-ML: A Machine Learning Framework for Exploring Mineral Relations and Classifying Common Igneous Minerals, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-13612, https://doi.org/10.5194/egusphere-egu23-13612, 2023.

The chemistry of erupted minerals and melts are commonly used to determine the pressures, temperatures and H₂O contents of magma storage regions beneath volcanic centres. In turn, these estimates are vital for hazard assessment, to understand the formation of critical metal deposits, and to inform models of continental crust formation. In the last few decades, more than 100 empirical and thermodynamic expressions have been calibrated using measurements of phases in experimental studies where these intensive parameters are known. By collating these different models into a computationally-efficient, open-source Python3 package, Thermobar, we can critically assess the performance of thermobarometers in igneous systems, and propagate analytical errors. When we apply published models for different mineral equilibrium to a new experimental dataset not used in model calibration, we find that stated errors vastly underestimate the true uncertainty when these workflows are applied to natural systems.

Specifically, we find that realistic calculation workflows involving Clinopyroxene (Cpx) equilibrium (e.g., iterating pressure and temperature) have uncertainties spanning the entire crust in most tectonic settings. Using Thermobar functions to propagate analytical error using Monte Carlo simulations, we suggest that these large errors result from imprecise analyses of minor elements such as Na in experimental (and natural) Cpx. Common analytical conditions used for Cpx yield highly correlated pressure-temperature arrays spanning the entire crust, which have been incorrectly interpreted as trancrustal storage in natural systems. Insuffucient analyses of each phase in experimental products means that this analytical error is not sufficiently mediated by averaging, so reported mineral compositions deviate from the true phase composition. This impacts thermobarometer calibration, as well as assessment of these methods using test experimental datasets.

Overall, we demonstrate that the development of Python3 infrastructure for common quantitative workflows in volcanology is vital to allow rigorous error assessment and model intercomparison; such assessments simply aren’t feasible using traditional tools (e.g., Excel workbooks). Specific changes to analytical, experimental and model calibration workflows (e.g., higher beam currents and count times in Na) will be essential to produce a more robust dataset to calibrate and test the next generation of more precise and accurate Cpx-based barometers. In turn, this will enable more rigorous investigation of magma storage geometries in a variety of tectonic settings (e.g., distinguishing true transcrustal storage vs. storage in discrete reservoirs).

How to cite: Wieser, P., Kent, A., Till, C., Petrelli, M., Wieser, E., Lubbers, J., Neave, D., Ozaydin, S., Donovan, J., Blatter, D., and Krawczynski, M.: Open-source Python3 tools for Thermobarometry: Revealing the good, the bad and the ugly of determining P-T-X conditions in igneous systems, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-3926, https://doi.org/10.5194/egusphere-egu23-3926, 2023.

The texture and morphology of igneous zircon indicates magmatic conditions during zircon crystallisation and can be used to constrain provenance. Zircons from porphyry copper deposits are typically prismatic, euhedral and strongly oscillatory zoned which may differentiate them from zircons associated with unmineralised igneous systems. Here, cathodoluminesence images of zircons from the Quellaveco porphyry copper district, Southern Peru, were collected to compare zircon textures between the unmineralised Yarabamba Batholith and the Quellaveco porphyry copper deposit. Quellaveco porphyry zircons are prismatic, euhedral and strongly oscillatory zoned, whereas the batholith zircons contain more variable morphologies and zoning patterns. We adopt a deep convolutional neural networks (CNNs) approach to demonstrate that a machine can classify porphyry zircons with a high success rate. We trial several existing CNN architectures to classify zircon images: LeNet-5, AlexNet and VGG, including a transfer learning approach where we used the weights of a VGG model pre-trained on the ImageNet dataset. The VGG model with transfer learning is the most effective approach, with accuracy and ROC-AUC scores of 0.86 and 0.93, indicating that a Quellaveco porphyry zircon CL image can be ranked higher than a batholith zircon with 93% probability. Visualising model layer outputs demonstrates that the CNN models can recognise crystal edges, zoning and mineral inclusions. We trial implementing trained CNN models as unsupervised feature extractors, which can empirically quantify crystal textures and morphology. Therefore, deep learning provides a powerful tool for the extraction of petrographic information from minerals which can be applied to constrain provenance in detrital studies.

How to cite: Nathwani, C., Wilkinson, J., Brownscombe, W., and John, C.: Mineral texture classification using deep convolutional neural networks: an application to zircons from porphyry copper deposits, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-16579, https://doi.org/10.5194/egusphere-egu23-16579, 2023.

The phrase “automated mineralogy” has been synonymous with electron microscopy in the geosciences for decades. The use of energy dispersive spectroscopy (EDS) to rapidly map samples and identify phases of interest has gradually seen a shift out of its original industry applications and into the academic research environment. A major issue for academics wishing to take advantage of this powerful tool is the original platforms are rigid in terms of their industry designed outputs, and there has been a lack of development in both the software and hardware capable of providing automated outputs.

However, rather than just looking backwards at what automated mineralogy was originally designed for, there is a more forward looking and important conversation as geoscience projects increase in scope. To think about geoscience in the context of topics such as big data, statistical relevance, and the use of increasingly prevalent machine learning techniques, we need to think about how we collect and store data. This naturally requires a greater integration of the problems geoscientists are trying to address with the solutions that microscopy and microanalysis equipment suppliers provide. Greater access to the data acquired on scientific equipment not only provides greater research flexibility but opens much smarter routes towards a future of correlated datasets with minimal user input.

Using quantitative chemistry as the basis for automated mineralogy provides unique capabilities for large area analysis such as thin sections. Quantitative textural information can be extracted from the sample such as grain sizes, shapes, and mineral associations, alongside quantitative geochemical data providing mineral classification, including mineral and whole rock/sample compositions. This provides a wealth of information for the petrologist to understand their sample and a one-stop-shop for many geoscience workflows. This is a ready made mechanism to generate large datasets across multiple samples in a consistent fashion – the key to big data.

Here we show one example of greater integration of data acquisition with user generated computational software showing the power of large area quantitative EDS mapping with geoscience-oriented functions of XMapTools. By importing calibrated, quantitative EDS maps XMapTools can be used to rapidly perform a variety of petrological calculations without the need for a separate, long-winded calibration step using microprobe data. Here we use quantitative EDS from high grade metamorphic rocks to obtain mineral and bulk compositions alongside textural information such as modal abundances. These mapped data are imported directly into XMapTools and can be used to generate oxide values, cation per formula unit (cpfu), end member proportions, and perform thermodynamic calculations.

How to cite: Taylor, R.: Reimagining automated mineralogy for the 21st century, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-7485, https://doi.org/10.5194/egusphere-egu23-7485, 2023.

Stream sediment geochemistry is a useful tool to derive geochemical insights into local geological units within stream sediment source areas. This has significant applicability within the field of mineral exploration where understanding regional geochemistry is fundamental to successful prospection and can facilitate the identification of critical metal deposits. This can help diversify the supply chain of critical metals, as well as tackle the deficit, especially for cobalt (Co). Cobalt is a growing component in many industrial processes but is mostly required for powering Li-Co batteries in Plug-in Hybrid Electric Vehicles (PHEV)¹. Demand for Co is growing exponentially in order to meet future carbon-neutral technological demand as part of joint UK-European initiatives towards a more environmentally sustainable society. 

The UK Geochemical Baseline Survey of the Environment (G-BASE) dataset is used to demonstrate that this technique provides a useful tool for isolating potential ‘Critical Minerals’² in host rocks across the UK Lake District, with priority targeting towards Co-bearing ores. We reduced the dimensionality of the G-BASE stream sediment data to create geochemical maps that identify a combination of volcanic, sedimentary, and plutonic lithologies lining up geological boundaries from established 50k scale geological maps of the area. This was conducted through a combined statistical and mapping approach within QGIS and ioGAS. The resultant lithogeochemical map of the region highlights the average geochemistry for each major lithological group with varying degrees of resolution.

This technique also allows for the identification of average ore metal concentrations (Ag, As, Bi, Co, Cu, Mo, Ni, Sn, Zn) for the Skiddaw Group and the Borrowdale Volcanic Group, two established host groups for As-Co-Cu-Ni mineralisation. Average concentrations of Co in the Skiddaw have been modelled to be 63.26 ppm, and in the Borrowdale volcanics to be 26.86 ppm. These values, combined with As, Cu, and Ni modelled concentrations, and other publicly available exploration-related data (structural maps, underlying batholith topography, mining history, mineral occurrences etc.) allowed us to identify 10 prospective areas of interest for possible As-Co-Cu-Ni mineralisation across these two lithological groups. Fieldwork was then undertaken to investigate several of these identified areas in order to establish the success of the model targeting tools. Ore metal-bearing minerals, mostly Cu-Fe-As phases, were identified both disseminated in local shales and andesites, and in hydrothermal quartz-chlorite veins at six sites investigated thus far. Characterisation of these minerals and host rocks is still in progress, making use of SEM-EDS and XRF analytics.

We demonstrate this workflow has strong applicability within critical metal exploration and should be applied in other, more prospective regions across the globe. The only pre-requisite to the mapping is the availability of stream sediment databases with sufficient resolution across target areas.

¹ Dehaine et al. 2021. BATCircle Project Report 04.
² Resilience for the Future: The UK’s critical minerals strategy, policy paper, 22^nd July 2022.

How to cite: Eskdale, A., Gough, A., and Johnson, S.: The applicability of stream sediment geochemistry as a combined geological mapping, and prospective exploration tool for As-Co-Cu-Ni mineralisation., EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-13447, https://doi.org/10.5194/egusphere-egu23-13447, 2023.

Here we introduce the AusGeochem platform enabling geoscientists to manage and publish sample based geoscience data in a consistent and FAIR way. 

The AusGeochem data platform has been built by the AuScope Geochemistry Network (AGN) and Lithodat Pty Ltd as part of a national digital infrastructure project to facilitate the sharing of data produced by geochemistry laboratories across Australia. 

In order to improve national geochemical facility data management, the AusGeochem platform uses detailed structured and standardized method specific data models for collating, preserving, and disseminating geochronology and isotopic data. 

AusGeochem provides researchers with a solution to record and utilize all of this information, while streamlining their workflow from collection to publication. Here we will give an overview of a typical sample workflow. 

The demonstrated workflow contains the following steps: Storing all sample details on-the-fly during sample collection with the field app. Managing the data by logging into the AusGeochem website. Adding subsequent geochemical analyses to the sample. Visualizing the data using analytic dashboards and graphs. Sharing the data with a team of collaborators. And finally, making the data referenceable by minting DOIs and IGSNs.

Manifold data is produced and captured along this process of collecting, analyzing and publishing. This is not confined to just the analytical results, but also a lot of meta information, such as involved people, instruments, funding sources, grant numbers, laboratories and institutions. Being rich in such metadata opens up the path to interesting new functionality, e.g. in the space of structured quantification and quality assessment of research projects.

The AusGeochem platform is of great help for a single researcher or a team of researchers by providing them with the means and tools to make their work more efficient and productive. However, the real benefit comes by having an access control layer, so data from multiple institutions can be  stored on one single platform. Since the data is now all stored in one detailed model, it gives researchers the possibility to apply analysis across data from all contributing institutions on the fly.  This means, no downloading of data from multiple sources required. And most importantly, no difficult and often impossible data preprocessing required.

How to cite: Theile, M., Noble, W., Beucher, R., McMillan, M., Boone, S., and Kohlmann, F.: From Field Application to Publication: An end-to-end Solution for FAIR Geoscience Data, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-11980, https://doi.org/10.5194/egusphere-egu22-11980, 2022.

How to cite: Theile, M., Noble, W., Beucher, R., Bedoya-Mejia, A., Boone, S., and Kohlmann, F.: From Fieldwork to Publication with AusGeochem: An Open end-to-end Solution for managing FAIR Sample Based Geoscience Data, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-4954, https://doi.org/10.5194/egusphere-egu23-4954, 2023.