GWGerhard Wörner

Gerhard Wörner

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Gerhard Wörner has a total of 109 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Multi-agent Geochemical Literature Data Mining System

    by Tianyu Yang, Karim Elezabawy, Daniel Kurzawe, Leander Kallas, Marie Traun, Bärbel Sarbas, Adrian Sturm, Stefan Möller-McNett, Matthias Willbold, Gerhard Wörner

    The increasing volume and complexity of geochemical literature pose major challenges for the sustainable curation of domain-specific databases such as GEOROC (Geochemistry of Rocks of the Oceans and Continents), the world’s largest repository of geochemical and isotopic data from igneous and metamorphic rocks and minerals, aggregating more than 41 million values from over 23,000 publications. Although GEOROC underpins a wide range of geoscientific research, the extraction and harmonization of metadata from publications still relies heavily on manual effort, which significantly limits the scalability.In this contribution, we present a novel information extraction architecture that moves beyond linear processing pipelines toward an Large Language Model (LLM)-based multi-agent system combining document layout analysis, schema-driven reasoning, and modality-aware extraction. Unlike generic LLM approaches that treat documents as continuous text streams, our architecture adopts a "Visual-First" strategy. We utilize a layout-aware backbone (MinerU, Niu et al., 2025) to decompose PDF manuscripts into a sequence of geometrically grounded primitive blocks, each representing a localized document region with associated visual and typographic features, preserving the geometric grounding essential for interpreting complex data tables. A routing agent subsequently validates and refines the initial layout classification, dynamically dispatching blocks to specialized downstream agents for text, table, or figure processing. This adaptive routing strategy improves robustness against layout variability across journals, publication years, and formatting styles.Central to the framework is an active schema agent that operationalizes the GEOROC metadata model. Rather than treating the database schema as a static template, this agent continuously provides extraction targets, normalization rules, unit standards, and conflict-resolution policies that guide all subsequent processing steps. Text blocks are handled by an Optical Character Recognition (OCR) driven information extraction agent, table blocks by a table parsing agent capable of reconstructing complex table structures, and figure blocks by a visual reasoning agent designed to interpret diagrams and digitize plotted values. Each agent produces structured candidate values enriched with confidence estimates and fine-grained provenance, including page-level and bounding-box references to the original document.The outputs of these modality-specific agents are consolidated by a merge-and-judge agent, which goes beyond simple aggregation. This agent performs cross-modal arbitration, unit harmonization, and deduplication, resolving conflicts between heterogeneous sources according to schema-defined priorities and data-quality criteria. The final result is a machine-readable JSON representation that preserves both extracted values and their evidential context.By combining layout grounding, adaptive routing, schema-driven reasoning, and judgment-based integration, this system delivers a robust and extensible approach to large-scale metadata extraction. The framework substantially supports the curation process and strengthens GEOROC’s role as a FAIR-compliant reference infrastructure by enabling more efficient reuse of published geochemical data in future geochemical research.References:Niu, J., Liu, Z., Gu, Z., Wang, B., Ouyang, L., Zhao, Z., ... & He, C. (2025). Mineru2. 5: A decoupled vision-language model for efficient high-resolution document parsing. arXiv preprint arXiv:2509.22186.

  • Open Access

    Interoperable Geochemical Data Infrastructures for Computational Magmatic Studies through Controlled Vocabularies

    by Leander Kallas, Marie Katrine Traun, Axel D. Renno, Dieter Garbe-Schönberg, Bärbel Sarbas, Adrian Sturm, Stefan Möller-McNett, Daniel Kurzawe, Matthias Willbold, Kerstin Lehnert, et al.

    Computational approaches in geochemistry are increasingly central to advancing our understanding of magmatic and volcanic systems as well as general Earth System processes. These methods rely on the integration of heterogeneous geochemical datasets spanning multiple spatial and temporal scales, analytical techniques, and material types. However, the effective reuse of such data remains limited by inconsistent metadata, ambiguous terminology, and insufficient interoperability between major geochemical data resources.The Digital Geochemical Data Infrastructure (DIGIS) addresses these challenges as part of the "OneGeochemistry Initiative" by modernizing and integrating two foundational geochemical databases: GEOROC (Geochemistry of Rocks of the Oceans and Continents) and GeoReM (Geological and Environmental Reference Materials). GEOROC and other databases provided to the community through the EarthChem Portal provide open access to millions of geochemical analyses of igneous and metamorphic rocks, minerals, and glasses, while GeoReM curates critically evaluated data on reference materials used for calibration, quality control, and uncertainty assessment in geoanalytical laboratories worldwide. Re-establishing and strengthening interoperability between these complementary resources is essential for computational studies that require traceable, reproducible, and quantitatively robust input data.This effort requires development and implementation of shared, machine-readable controlled vocabularies covering sample descriptions, lithology, mineralogy, geological setting, analytes, material matrices, methods, and reference materials. These vocabularies harmonize legacy data in GEOROC and GeoReM, while remaining compatible with international data standards developed by the OneGeochemistry Initiative. By linking observational data and rich metadata, the integrated system enables more flexible data filtering, uncertainty-aware model input, and reproducible benchmarking of computational results.Recent computational studies illustrate the scientific value of such harmonized geochemical data infrastructures. Machine-learning approaches have successfully leveraged large global GEOROC data compilations to quantitatively discriminate tectono-magmatic settings and extract compositional features related to magma generation and evolution. Combining volcanic eruption histories with interoperable GEOROC and PetDB datasets from the EarthChem portal has further enabled data-driven exploration of magma compositional variability across tectonic environments. In parallel, emerging machine-learning-based petrological models, such as thermobarometers trained on large, standardized compositional melt and mineral datasets, demonstrate how consistent geochemical input data are critical for inferring magma storage conditions and differentiation.This contribution highlights how sustained investment in FAIR-aligned geochemical data infrastructures directly support advances in computational magmatic studies. By improving interoperability of international geochemical databases, such as GEOROC and GeoReM, through controlled vocabularies, we provide a foundation for computational volcanic and magmatic studies, uncertainty-aware analysis, and quantitative modelling.

  • Open Access

    How comparable are geochemical datasets really and why it matters

    by Marie Katrine Traun, Axel D. Renno, Leander Kallas, Matthias Willbold, Tod Waight, Dieter Garbe-Schönberg, Gerhard Wörner

    In geochemistry, big data applications at both global and regional scales rely on compilation and aggregation of numerous smaller datasets to form a big dataset. For this reason, geochemical data compilations are vulnerable to systematic biases between the smaller datasets they are composed of. These biases relate to the analytical methods and procedures, laboratories, instruments, sample preparation, detection limits, mass interferences, sample matrices, etc. In other words, the preparation and analysis of each sample batch is unique beyond the “method” or “laboratory” bias. This uniqueness and potential offsets it might cause between analytical batches we define as inter-study bias. The only quantitative way to evaluate inter-study bias in geochemical data compilations is through metadata and specifically assessment of analytical data of geochemical reference materials. Reference materials are substances of known composition measured alongside unknown samples, as is a standard good practice during routine geochemical analyses. In an ideal world, all geochemical studies report analyses and values of reference materials and analytical methods and analyses have been refined and calibrated to match the reference material’s certified value within uncertainty. Only in this case can the inter-study bias be considered negligible. Accordingly, most geochemical big data compilations are based on this assumption and do not explicitly assess the metadata for potential inter-study bias. In the real world, perfectly calibrated analyses are often not the case and metadata uncommonly reported.To assess the comparability, compilability and inter-study bias between geochemical datasets, we have developed several data quality and outlier-detection tools based on the Geological and Environmental Reference Material database - GeoReM. We use these tools to showcase the implications of inter-study bias for global geochemical interpretation models using two well-known geochemical big data research topics: 1) identification of compositional end-members for oceanic basalts and the origin of their source mantle components (colloquially called “the mantle zoo”) and 2) compositional signatures of zircons as tracers for the growth, reworking and evolution of the continental crust. Our take home message: Geochemical datasets must be comparable to be compilable. We therefore advocate the assessment of your inter-study bias as well as comprehensive reporting of your metadata and reference materials, so that computational geochemistry can progress as a subdiscipline of big data science.Keywords: Reference material, GeoReM, outlier, metadata, method bias, isotopes, mantle geochemistry, zircon, GEOROC, crustal evolution

  • Open Access

    Significance and Future of Data Infrastructures for the Geochemical Research Community

    by Gerhard Wörner, Marthe Klöcking, Kerstin Lehnert, Kirsten Elger, Digis-Georoc Team

    The GEOROC and PetDB databases provide peer-reviewed geochemical data on igneous rocks, minerals and related materials for >25 years to cover the full range of igneous compositions, mantle xenoliths and minerals. Combined they provide access to more than 48.2 million individual data values from around 27,000 publications through web applications for search, filtering and download. These comprehensive datasets support large-scale regional and global geochemical data-based research spanning traditional geochemical studies to data-driven and machine-learning approaches.GEOROC’s holdings have reached over 40.8 million data values from more than 23,000 publications focusing on ocean islands, continents and subduction zones. The PetDB database complements with ca. 7.4 million data values for igneous and metamorphic samples of the ocean floor, ophiolites, mantle xenoliths, tephra, and arc rocks.The DIGIS project is modernizing the GEOROC data infrastructure in alignment with FAIR principles by introducing a new API, improved web interface, and unified data model. Further, topical global collections of data are extracted into individual DOI-minted data sets that are regularly updated from the GEOROC data holdings. These compilations and additional author-contributed data sets with rich metadata are accessed through GFZ Data-Services. GEOROC has recently been reconnected with the updated GeoReM database on geochemical reference materials. PetDB is part of the EarthChem data services and the IEDA2 data facility. PetDB was migrated to a new architecture and a new, simplified search interface was released in 2025 to improve usability. EarthChem also offers repository services where researchers can publish and archive their data.Based on close collaborations between PetDB and GEOROC, the EarthChem Portal has provided for nearly 20 years a central access point to the content of both databases, as well as several smaller databases. Today, nearly 50 million data values are accessible at the ECP.While the EPOS data resources are strong on geophysical (and other types of) data, EPOS has lacked a systematic inclusion of geochemical data from rocks on the European continent. The data services that the geochemical research community provide on geochemical compositions of rocks minerals and ore deposits globally has the potential to become a strong contribution to the EPOS data platform. To this end, we offer collaboration with EPOS to provide access points for two types of geoscience data: curated geochemical data on rocks and minerals in a domain-specific data base and large compiled selected data sets on specific types of rocks and minerals and/or from specific geological or geographic settings in the DIGIS-GEOROC repository at GFZ-data services. This also requires further developments: Under the umbrella of OneGeochemistry and NFDI4Earth, DIGIS, EarthChem and other initiatives such as the Australian Geochemistry Network are developing authoritative vocabularies and metadata standards, as well as interoperability and integration across different global geochemical databases. Further, together we develop tools for data quality assessment for improved data usability. These advances also broaden the applicability of geochemical data beyond hardrock oriented research to fields such as environmental science, archaeology and geohealth, demonstrating how FAIR-aligned geochemical infrastructures enhance reproducible research in Earth System Science and interdisciplinary collaboration.

  • Open Access

    Advancing FAIR geochemical data: 25 Years of GEOROC database service

    by Leander Kallas, Marthe Klöcking, Kirsten Elger, Bärbel Sarbas, Adrian Sturm, Stefan Möller-McNett, Matthias Willbold, Gerhard Wörner

    The GEOROC synthesis database, a pioneering open-access resource for geochemical and isotopic data, marks 25 years of service to the geoscience community. Over its history, GEOROC has compiled data from more than 22,750 publications in the field of geochemistry, and provides free access to over 39 million individual data values, primarily on igneous and metamorphic rocks, minerals and their inclusions. As a cornerstone for interdisciplinary research, GEOROC is complementary to other geochemical synthesis databases like PetDB, AstroMat and GeoReM, in facilitating reuse of data for innovative studies that leverage data analytics and machine-learning approaches across geoscientific disciplines and beyond.The Digital Geochemical Data Infrastructure (DIGIS) project for GEOROC 2.0 is providing an up-to-date IT infrastructure that aligns GEOROC with the FAIR principles. Data findability and accessibility are ensured through the newly developed API and the improved GEOROC web interface that allows users to retrieve a variety of distinct data products and services, including a fully customizable search functionality. Interoperability is achieved via implementation of a feature-based data model compatible with the OGC Observations and Measurements standard and controlled, machine-readable vocabularies that harmonize geospatial, analytical and sample-related metadata, and enabling seamless integration in multiple databases and portals (e.g., EarthChem). Reusability is further supported by archiving time-stamped GEOROC data products in the DIGIS Data Repository, hosted by GFZ Data Services, where datasets with digital object identifiers (DOIs) are archived for the long-term. Additionally, researchers are encouraged to directly submit new or already “published” datasets to this domain repository—through standardized (meta-)data templates, ensuring high-quality data submissions that facilitate data quality assessment and reuse.In collaboration with national and global initiatives, such as OneGeochemistry and NFDI4Earth, the DIGIS project further promotes practical approaches to the FAIR principles for geochemistry by developing unified controlled vocabularies for geochemical data and their metadata (e.g., analytical methods, sample description, location). These vocabularies also integrate external standards, such as the International Mineralogical Association’s "List of Minerals" and MinDat’s "Subdivisions of Rocks," alongside newly developed (and published) frameworks for categories such as geological setting and analytical methods (collaboration with EarthChem). By harmonizing metadata across geospatial, analytical and sample-related categories, these efforts ensure consistency, improve data quality assessment and control and enhance interoperability across data systems, including but not limited to GEOROC, PetDB, and AusGeochem. Such advancements expand the potential applications of geochemical data, fostering innovation in fields such as environmental science, remote sensing, archaeology and geohealth.With 25 years of experience and ongoing innovation through the DIGIS project, the GEOROC database exemplifies how operationalizing the FAIR principles enhances its value as a critical resource for the geoscience community. By providing both FAIR and open data, GEOROC empowers researchers to conduct reproducible, impactful studies and fosters interdisciplinary collaboration, driving innovation and advancing progress across the geosciences.

  • Open Access

    Unified Vocabularies for Geo- and Cosmochemical Data Systems

    by Leander Kallas, Marthe Klöcking, Lucia Profeta, Stephen Richard, Annika Johansson, Kerstin Lehnert, Manja Luzi-Helbing, Bärbel Sarbas, Hannah Sweets, Dieter Garbe-Schönberg, et al.

    Global compilations of geo- and cosmochemical data are increasingly leveraged to address exciting new research questions through data-analytics and machine-learning approaches. These invaluable datasets are maintained and made accessible as synthesis databases, such as GEOROC and PetDB catering to terrestrial igneous and metamorphic rocks; AstroMat Data Synthesis encompassing diverse astromaterial samples; and GeoReM a comprehensive resource for geochemical, environmental and biological reference materials. The GEOROC and PetDB databases for igneous and metamorphic rocks collectively aggregate data from thousands of publications, combining over 42 million single data values (major and trace elements, stable and radiogenic isotope ratios, radiometric ages) for bulk rock, glass, as well as minerals and their inclusions.The diverse focus of these data systems include data from different sources and metadata makes data integration and interoperability challenging. The DIGIS and EarthChem projects are working towards designing machine-readable unified vocabularies for their data systems to achieve full interoperability. These vocabularies, associated with primary chemical data as well as geospatial, analytical and sample metadata, encompass many categories describing geographic location, sampling technique, lithology and mineral types, geological and tectonic setting, as well as analytes, analytical methods, reference materials, and more.Wherever possible, external machine- and/or human-readable external vocabularies from respected authorities are incorporated, such as MinDat’s "Subdivisions of Rock," the International Mineralogical Association’s "List of Minerals" (Warr, 2021), and the International Union of Pure and Applied Chemistry’s chemical terminologies. For remaining categories, a set of local vocabularies are developed by our group (e.g. analytical methods, see Richard et al. 2023). The collaborative effort between DIGIS, EarthChem, and the Astromaterials Data System is leading to an advanced vocabulary ecosystem relating samples, data, and analytical methods in geo- and cosmochemical research that reaches from local- to community-driven and, eventually global connections.Establishing a globally accepted vocabulary not only contributes to building interoperability between our existing geo-and cosmochemistry synthesis databases, but will also help pave the way toward interoperability with the GeoReM database, linking data with analytical methods and reference materials to provide means for data quality control and assessment of analytical uncertainty.Finally, the unified vocabularies of EarthChem, GEOROC, and GeoReM will advance the creation of a global network of geochemical data systems as promoted by the OneGeochemistry initiative (Klöcking et al., 2023; Prent et al. 2022), connecting and integrating the broadest range of geoanalytical data generated, for example, in studies of environmental samples, archeological artefacts, or geohealth matters.We report on these goals, achievements, state of advance, and challenges and seek community engagement and feedback. ReferencesKlöcking, M. et al. (2023). Community recommendations for geochemical data, services and analytical capabilities in the 21st century. In Geochimica et Cosmochimica Acta (Vol. 351, pp. 192–205).Prent, A. et al. (2023) Innovating and Networking Global Geochemical Data Resources Through OneGeochemistry. Elements 19, Issue 3, pp. 136–137.Richard, S. et al. (2023) Analytical Methods for Geochemistry and Cosmochemistry. Concept Scheme for Analysis Methods in Geo- and Cosmochemistry. Research Vocabularies Australia.Warr, L. N. (2021). IMA–CNMNC approved mineral symbols. Mineralogical Magazine, 85(3), 291-320.

  • Open Access

    GEOROC 2.0: A Globally Connected Geochemical Database to Facilitate Interdisciplinary, Data-Driven Research

    by Gerhard Wörner, Marthe Klöcking, Adrian Sturm, Bärbel Sarbas, Leander Kallas, Stefan Möller-McNett, Kirsten Elger, Daniel Kurzawe, Matthias Willbold

    The GEOROC database is a leading, open-access source of geochemical and isotopic datasets that provides access to curated compilations of igneous and metamorphic rock and mineral compositions from >20,600 publications. It is a data resource that supports and facilitates hundreds of new research publications each year across multiple geoscientific and related disciplines.This presentation is to “advertise” to the geochemical community this data product and our ongoing efforts to improve the service by providing FAIR (findable, accessible, interoperable and reusable) geochemical data. We will also describe some recently published research where authors were using large geochemical data compilations such as GEOROC and PetDB for innovative approaches in digital geochemistry.To further support such research also in the future, the Digital Geochemical Data Infrastructure (DIGIS) initiative is developing a new IT and data infrastructure for GEOROC 2.0 to enable modern solutions to data submission, discovery and access. GEOROC data compilations are made accessible via a web search interface and an API. In addition, DIGIS maintains a direct data pipeline between the data compiled in GEOROC and the EarthChem Portal. Hence, GEOROC represents one of six different geochemical databases that can be queried and accessed synchronously within the EarthChem Portal. The DIGIS infrastructure further partners with GFZ Data Services, a domain repository for geosciences data, hosted at GFZ, offering data publication services with assigned digital object identifiers (DOI). Individual researchers can directly submit their geochemical datasets to the repository (using the EarthChem Data Templates) where they are archived for the long term. Regular thematic snapshots of the GEOROC synthesis database are archived in the GRO.data repository of the University of Göttingen.Part of this cooperation is the development of standardised vocabularies and data reporting to enhance interoperability of geo- and cosmochemical data systems. Harmonized data entry for the GEOROC, PetDB and Astromat synthesis databases will avoid duplication and ensure consistent data and metadata. With these efforts, and as a participant of the OneGeochemistry(1,2) initiative, DIGIS is working towards the goal of globally harmonised geochemical data to enable interdisciplinary, data-driven research. ReferencesKlöcking, M. et al. (2023). Community recommendations for geochemical data, services and analytical capabilities in the 21st century. In Geochimica et Cosmochimica Acta (Vol. 351, pp. 192–205).Prent, A. et al. (2023) Innovating and Networking Global Geochemical Data Resources Through OneGeochemistry. Elements 19, Issue 3, pp. 136–137.


  • Open Access

    Magma storage conditions of Lascar andesites, central volcanic zone, Chile

    by André Stechern, Magdalena Blum-Oeste, Roman E. Botcharnikov, François Holtz, Gerhard Wörner

    Lascar volcano, located in northern Chile, is among the most active volcanoes of the Andean Central Volcanic Zone (CVZ). Its activity culminated in the last major explosive eruption in April 1993. Lascar andesites which erupted in April 1993 have a phase assemblage composed of plagioclase, clinopyroxene, orthopyroxene, Fe–Ti oxides, and rhyolitic glass. To better constrain storage conditions and mechanisms of magmatic differentiation for andesitic magmas in a thick continental crust, crystallization experiments were performed in internally heated pressure vessels at 300 and 500 MPa, in the temperature (T) range of 900–1050 °C, at various water activities (aH2O) and oxygen fugacities (logfO2 between QFM+1.5 and QFM+3.3 at aH2O =1; QFM is quartz–fayalite–magnetite oxygen buffer). The comparison of experimental products with natural phase assemblages, phase compositions, and whole-rock compositions was used to estimate magma storage conditions and to reconstruct the magma plumbing system. We estimate that Lascar two-pyroxene andesitic magmas were stored at 975±25 °C, 300±50 MPa, and logfO2 of QFM+1.5±0.5 , under H2O-undersaturated conditions with 2.5 wt % to 4.5 wt % H2O in the melt. The geochemical characteristics of the entire suite of Lascar volcanics indicates that a fractionating magmatic system located at a depth of 10–13 km is periodically replenished with less evolved magma. Some eruptive stages were dominated by volcanic products resulting most probably from the mixing of a mafic andesitic magma with a felsic component, whereas compositional variations in other eruptive stages are better explained by crystal fractionation processes. The relative importance of these two mechanisms (mixing vs. crystal fractionation) may be related to the amount and frequency of magma recharge in the reservoir.


  • Open Access

    Molybdenum Isotope Systematics of the Kamchatka Subduction Zone System

    by Matthias Willbold, Gerhard Wörner

    Molybdenum (Mo) isotopes in magmatic rocks are a promising tool in high-temperature isotope geochemistry. In particular, basalts from subduction zones that are geochemically controlled by mass transfer through slab-fluid addition have systematically higher δ98Mo values (i.e. measured 98Mo/95Mo ratio in a sample relative to that in a standard) than the depleted mantle (δ98Mo = –0.21‰). In these rocks, the elevated δ98Mo values are linked to high Pb/Ce and high (238U/230Th) ratios and can be reconciled by the addition of isotopically heavy Mo via a slab fluid component1,2. So far, these systematics are best expressed in subduction zone basalts from the Mariana and Izu arc systems that also form coherent mixing trends between fluid-enriched mantle domains in δ98Mo versus 143Nd/144Nd and 176Hf/177Hf space1,2.The Kamchatka arc system represents the northernmost expression of the W-Pacific convergent margin. Volcanic front lavas are dominated by slab-to-mantle mass transfer through fluid transport, whereas subduction of the Emperor seamount ridge gives rise to back-arc basalts with a geochemical and isotopic affinity to within-plate basaltic rocks3.Here, we report δ98Mo data for 47 basalts from an E-W transect across the Kamchatka peninsula that have previously been analysed for their major, trace element, radiogenic and stable isotope data. The δ98Mo data extent the trend defined by samples from the Marianas and Izu arcs starting from moderately high δ98Mo and Pb/Ce values towards sub-depleted mantle δ98Mo and mantle-like Pb/Ce ratios that indicate the presence of a source component formed by partial melts of a rutile-bearing mafic crust4.The common geochemical and isotopic trends formed by the combined Mariana – Izu – Kamchatka datasets suggest a surprisingly uniform Mo isotope composition of a subduction zone fluid endmember for more than 5000 km along-strike of the Circum-Pacific subduction zone system. Our data also confirm the presence of an enriched source component in the Kamchatka mantle wedge, possibly originating from the subducted Emperor seamount chain5.1Freymuth, H., et al., EPSL 432, 176-186 (2015). 2Villalobos-Orchard, J., et al., GCA 288, 68-82 (2020). 3Churikova, T., et al. JPet 42, 1567-1593 (2001). 4Chen, S., et al., Nat. Comm. 10, 4773 (2019). 5Shu,Y., et al.,Nat. Comm. 13, 4467 (2022).