This was the FAIRagro Community Summit 2026

This keynote highlights the decisions and tasks involved in setting up the FAIRagro technical service platform. Spotlighting examples, we take the audience behind the scenes and provide insights into the search for optimal solutions for technical services that support experimental science from conducting experiments to data collection and analysis.

More about Matthias Lange

Agricultural research data is often scattered across discipline-specific repositories, making discovery difficult and time-consuming. The FAIRagro Search Hub solves this by serving as a cross-disciplinary search portal that connects researchers to datasets and repositories relevant to agrosystems. Based on metadata harvested from infrastructures like re3data.org and FAIRagro’s partner repositories, it creates a harmonized overview without duplicating data storage. This session offers a practical demonstration of the Search Hub’s features and contents that it makes available to you. We invite you to test the service yourself and share your insights to improve its usability and coverage.

More about Julian Schneider & Ataul Haleem
Contributors: Julian Schneider & Ataul Haleem

High-quality metadata is crucial to make datasets findable and interoperable. While Research Data Management (RDM) continues to develop new metadata schemas, leading to processes generating more consistent and domain-specific metadata, legacy metadata still remains in its original state. Unstructured information, such as abstracts or titles, can contain the relevant information, but identifying and extracting it needs to be (semi-)automated to be feasible.

The presented case study applies trained AI agriculture models on FAIRagro Search Hub contents for automated extraction and metadata enrichment, building on the results of FAIRagro Pilot Use Case 8, “Increasing FAIRness of FAIRagro data through AI supported metadata enrichment”. It showcases the extraction results and their representation. It then evaluates which new search queries are possible through the additional, Agrischemas-based metadata and how these automatically generated contents can be presented to users transparently, enabling them to come to informed decisions.

Contributors: Gabriel Schneider & Julian Schneider

The German Weather Service (DWD) offers a wide range of meteorological and climate data free of charge and updated daily through its OpenData platform (opendata.dwd.de), which metadata are maintained via “Geonetwork”. These data are highly relevant for scientific research, commercial applications, and public services. By integrating this (meta) data into the Geoportal Germany (GDI‑DE), they become accessible to a larger user base and improve findability and interoperability.

GDI‑DE serves as a central hub for geodata-based information in Germany and enables standardized provision and use of geodata. Through this integration, DWD’s data will be also made available in the FAIRagro repository.

The combination of DWD’s OpenData platform, availability via GDI‑DE, and integration into FAIRagro repositories ensures that meteorological and climate data are not only easily accessible but also meet the highest standards for data processing and use. This underscores the importance of data infrastructure for the scientific community and agricultural practice.

More about Rafael Posada

Planning on how to manage its data is an important part of every research project. Data Management Plans (DMPs) offer researchers a guided way to identify, evaluate, and store their requirements collaboratively by filling out predefined templates.
FAIRagros DMP service, based on the Research Data Management Organizer (RDMO), enables researchers to create DMPs for their research projects based on templates of common funders, such as DFG and Horizon Europe. Additionally, a customized template for agrosystem sciences has been developed. It supports researchers with domain-specific answers and help texts as well as links to other FAIRagro services. In this regard, the FAIRagro Helpdesk operated by the Data Steward Service Center (DSSC) offers tailored mentoring and hands-on support for researchers in the agricultural sciences, integrating the DMP process closely with best practices in research data management and beyond.
The live demo showcases RDMOs functionalities and how researchers can use it to improve their daily research data management practices, presenting how it is embedded within FAIRagros service landscape. Furthermore, the Community Summit acts as a starting date for public access to FAIRagros RDMO instance, with a developed domain-specific template as an example.

More about Antonia Leidel
Contributors: Gabriel Schneider, Elena Rey Mazón

As an output from a FAIRagro workshop, specifically designed for the ressort of food, agriculture, forestry and tourism in 2024, we set out to generate a DMP questionaire for RDMO. During the development, we included five governmental research institutions, IT-departments as well as responsible departments for datasecurity and archives. Since June 2025, this questionaire is relevant for funding as well as for internal, institutional usage. Here, we talk about our experiences during this collaborative process.

As part of FAIRagro, several data quality–related services have been introduced, each targeting different stages of the data life cycle. This presentation will highlight the distinct roles of these services for the community and illustrate how they collectively contribute to the bigger picture of data quality. We will also discuss the potential of linking these services to enhance their overall impact within the further course of the FAIRagro project.

More about Mahdi Hedayat Mahmoudi
Contributors: Sven Gedicke, Susanne Lachmuth, Markus Möller, Jascha Jung, Gabriel Schneider, Mahdi Mahmoudi

The ability to create FAIR computational workflows is increasingly critical for ensuring transparency, reproducibility, and reusability in data-driven research. By using computational workflows, researchers can automate long sequences of scripts and combine tools from different programming languages into fully managed, reproducible pipelines, making FAIR practices easier to implement without sacrificing flexibility. To support this, the Scientific Workflow Infrastructure (SciWIn) provides complementary clients: SciWIn-Client and SciWIn-Studio. Those applications are designed to support efficient interaction with computational workflows based on the Common Workflow Language (CWL).

Both applications address the same core requirement: the easy, fast, and frictionless creation of FAIR computational workflows, while targeting different user groups. The command-line application SciWIn-Client (s4n) is aimed at users who are comfortable working at the command line and seek automation and scripting capabilities. In contrast, the upcoming SciWIn-Studio provides a graphical user interface for users who prefer visual interaction and guided workflow construction.

This presentation demonstrates how both tools can be used individually and in combination to create a comprehensive automated pipeline, covering the full process from measurement to publication-ready figures.

More about Jens Krumsieck
Contributors: Measure 4.4

We present a tool that enables large-scale comparisons of agricultural field polygons across datasets with differing granularity, quality, and temporal reference. By building spatial correspondences between the datasets under comparison, our approach supports spatial and semantic quality validation, consistency analyses, data fusion, and the analysis of temporal changes. Our web-based visualization tool allows users to visually explore different comparison metrics across the datasets. Implemented as an interactive map viewer, it provides visualizations specifically designed to effectively communicate spatial and semantic alignment in a geographically contextualized manner.

More about Sven Gedicke
Contributors: Sven Gedicke, Jan-Henrik Haunert

This contribution presents an interactive FAIRagro tool implementing the Application-Specific Data Quality Matrix (AS-DQM) developed in Measure 3.3 to support application-oriented assessment of data quality and fitness-for-use (DFFP). Building on FAIRagro Report (https://zenodo.org/records/17981173), the approach demonstrates how quality-relevant information contained in scientific publications and dataset documentation can be systematically extracted and formalized.

The live demo showcases a Streamlit-based pipeline that ingests textual documents and transforms narrative descriptions into a standardized, machine-actionable AS-DQM JSON representation, including application context, validation strategies, uncertainty information, limitations, and explicit suitability statements aligned with ISO 19157.

Using a Germany-wide phenology dataset as a pilot example, the demo illustrates how dataset-level quality documentation can be linked to application-specific requirements, making assumptions and constraints explicit and supporting informed reuse decisions within FAIRagro services.

More about Mahdi Hedayat Mahmoudi
Contributors: Mahdi Hedayat Mahmoudi, Markus Möller(TA3/M3.3)

FAIRagro’s Authentication and Authorization Infrastructure provides secure single sign-on across FAIRagro tools and platforms. For researchers and service providers, this means less effort managing accounts, easier collaboration across institutions, and faster access to data and services. Shared identities, roles, and groups support controlled access while keeping workflows simple and transparent. By offering a common, reliable login for all, FAIRagro strengthens its community, lowers barriers to participation, and allows users to focus on research rather than access management.

More about Carmen Scheuner
Contributors: TA4

Metadata as data about data: We talk about it’s importance for beeing able to find data, to understand data, to make data machine-readable, citable, reusable. We discuss metadata standards, the semantic web, controlled vocabularies and ontologies. But, practically speaking, where and how to start? How does broad agricultural researcher know what metadata to create, how to collect, where to store and structure it during the research project?

This practical guide to collecting metadata in agricultural research aims to provide an easy and practice-oriented introduction to the world of data documentation. A decision tree guides researchers with simple, practical questions through the topics:

1. Why data should be documented;
2. Which forms and formats are appropriate for data documentation in your own project;
3. Which metadata should be collected for your own research data;
4. How the metadata can be created;
5. Which forms of storage and linking with the research data should be considered; and
6. How the quality of the metadata can be ensured.

In addition to the decision tree, this document also contains brief explanations of terms  that should be understood in order to answer the questions in the decision tree meaningfully. In addition, a metadata guide with many agricultural science examples provides an overview of metadata, metadata schemas, terminologies and various forms of metadata collection and storage. Internal links in the decision tree refer to specific term explanations or further explanations in the metadata guide. 

More about Sophie Boße
Contributors: TA2, TA3: M 2.2, M2.4, M2.5, M3.2, Anne Sennhenn, Elena Rey Mazón, Lucia Vedder, Gabriel Schneider

A major challenge in research data management (RDM) training is effectively and meaningfully reaching the target audience. Many researchers lack the time or motivation to engage with generic RDM training because they may perceive it as too abstract or not applicable to their needs.

To reach our community and provide training that addresses their everyday needs, we collaborate with universities like our associated partner, the University of Kassel, and research institutes like the FZ Jülich. Within this cooperation, M2.4 and M2.5 work together to provide discipline-specific RDM content, while the cooperation partners provide information on institutional infrastructures and services. Additionally, drawing on the experience of our UC5, we can provide hands-on examples of real-world researcher problems and solutions within the research data management journey. 

More about Sophie Boße
Contributors: TA2, UC5, University of Kassel (associated partner), Lucia Vedder, Lars Grygosch, Ireneusz Kleppert, Sabrina Jordan

Background: Project funding requires data management plan.
Issue: Funder‘s guidelines gap considering DMP (Fig. 2).
Approach: RDMO allow to map variable guidelines.

Method: In 2025, BMLEH‘s Federal Research Centre (RFE) launch an initiative to harmonize DMP templates in BMLEH‘s funding agencies FNR (Fachagentur Nachwachsende Rohstoffe e. V.) and BLE (Bundesanstalt für Landwirtschaft und Ernährung).

Result: In December 2025 FNR and BLE agreed to provide a minimal, common template on data management concerns. The DMP template address especially funding agency relevant content.

Outlook: In future, DMP templates are open available in research data management tools, funding agencies receive the requested DMP via API. BMLEH’s funding agencies and RFE plan to continue their cooperation to improve DMP’s.

More about Torsten Thalheim
Contributors: TA2

The EmiMod research project (“Advancing Techniques for Measurement, Modeling, and Evaluation of Emissions in Livestock Buildings”) represents a co-operation of 9 research institutions across Germany and encompasses 12 work packages. The project focuses on advancing methodologies to determine emission rates from livestock farming, with particular emphasis on improving data acquisition in animal-welfare-oriented barns with natural ventilation and outdoor yards. Targeted emissions include ammonia, greenhouse gases, odours, and bioaerosols. The primary objective is to simplify investigation methods and to develop adaptable assessment procedures that can be applied across diverse research aims.
 
KTBL (Kuratorium für Technik und Bauwesen in der Landwirtschaft e.V.) is responsible for project coordination as well as research data management. Within the project’s research data management framework, all partners are required to create data management plans (DMPs) documenting the handling and lifecycle of the data collected in their work packages. Considering the complexity of the involved organizations and the different work packages, we decided to employ FAIRagros RDMO based service as a tool. In this contribution, we describe the process of creating the project’s DMP template and its implementation in RDMO. Its structured, FAIR-aligned framework has proven highly effective in harmonizing procedures, supporting high-quality documentation in a less time-consuming and user-friendly way, and promoting interoperability across institutions. Furthermore, we outline the current data management workflow established in the project and illustrate how the RDMO tool supports this process.

DiPredict is an associated project within DiP Saxony-Anhalt Model Region of the Bioeconomy – Digitalization of Plant Value Chains. The aim of the DiP project DiPredict is to promote digitalization in agriculture through AI-based optimization of selection processes under drought stress in wheat breeding.
As one of 20 DiP projects, DiPredict is representative of the challenges posed by multimodal and multiple data sources in research within the DiP context, as well as in plant and agricultural research in general. The project focuses on the development of a workflow for integrating diverse data sources and heterogeneous data points in order to support data-driven research and decision-making processes.

Base4NFDI is a cross-consortium initiative within Germany’s National Research Data Infrastructure (NFDI) that develops foundational, interoperable components to support FAIR-aligned research data management across disciplines. As NFDI consortia such as FAIRagro work to address domain-specific requirements in agrosystems research, Base4NFDI provides shared, reusable infrastructure that reduces duplication of effort and promotes interoperability throughout the broader ecosystem.
This poster introduces the mission, development approach, and collaborative processes of Base4NFDI, with a focus on how consortia can engage with and benefit from the emerging portfolio of foundational components. Particular attention is given to mechanisms for requirement gathering, integration pathways, and community feedback loops that ensure solutions remain applicable across heterogeneous research domains.
The poster also outlines how Base4NFDI supports alignment between local RDM practices and NFDI-wide standards, enabling smoother technical integration and more consistent FAIR workflows. By contributing requirements, testing components, and participating in shared development processes, FAIRagro community members can help shape infrastructure that meets both cross-domain and agriculture-specific needs.
The aim is to highlight opportunities for collaboration between FAIRagro and Base4NFDI, fostering a coherent and sustainable research data landscape within NFDI.

The increasing volume, heterogeneity, and complexity of life science and agrosystem data require robust infrastructures and harmonised Research Data Management (RDM) practices to ensure Findability, Accessibility, Interoperability, and Reusability (FAIR). The German Network for Bioinformatics Infrastructure (de.NBI) and ELIXIR Germany address these challenges by providing domain-relevant bioinformatics services, standardised training programmes, and scalable compute resources through the de.NBI Cloud. These components support reproducible data processing, high-throughput analyses, and the implementation of structured metadata and ontological standards.
These efforts closely align with the aims of the FAIRagro consortium, which is building a FAIR research data management ecosystem for agrosystem science. Synergies arise from shared challenges such as integrating heterogeneous datasets, adopting community standards, and providing scalable computation for data-intensive workflows. Additional alignment emerges through joint training activities, knowledge exchange, and collaborative contributions to interoperable workflows and metadata practices. By connecting infrastructures and expertise across domains, solutions developed within FAIRagro can be linked seamlessly to the wider national and European life science data landscape fostered by de.NBI & ELIXIR Germany.
This poster showcases how de.NBI & ELIXIR Germany support researchers through training, cloud compute services, and practical RDM resources, and how these efforts complement FAIRagro’s mission. By strengthening cross-domain collaboration and promoting FAIR-aligned practices, both networks contribute to a sustainable ecosystem that enhances data quality, interoperability, and reuse across life science and agrosystem research communities.

Research Data Management (RDM) has become a core component of responsible, transparent, and reusable research. Despite growing policy and infrastructure investments, RDM practices often remain fragmented, with misalignment between institutional strategies and the everyday workflows of researchers and community members. We present ARC as an enabling Research Data Management framework designed to support both institutions and community members through an integrated, scalable, and practice-oriented approach.
ARC bridges individual research practices with institutional and community-level infrastructures by providing a coherent framework that spans the full data lifecycle, from data planning and creation to curation, sharing, and long-term preservation. For community members and individual researchers, ARC emphasizes usability, lightweight adoption, and flexibility, enabling good RDM practices without imposing excessive technical or administrative burdens. For institutions, ARC supports harmonization across services, policies, and standards, enabling coordination while respecting disciplinary diversity and local governance structures.

By explicitly connecting bottom-up practices with top-down institutional support, ARC enables sustainable RDM ecosystems that foster collaboration, interoperability, and reuse. The framework promotes shared ownership of data stewardship responsibilities, reducing duplication of effort and strengthening collective capacity. Rather than treating RDM as a compliance-driven obligation, ARC frames it as a socio-technical practice embedded in research cultures and communities.

As research becomes increasingly collaborative and data-intensive, ARC offers a pathway toward coherent, enabling, and community-aligned RDM practices.

This presentaiton will highlight the opportunities of the applied FDO concept for the federated FAIRagro network and the use cases.

More about Daniel Arend
Contributors: Tobias Wamhof, Timo Mühlhaus, Kevin Schneider

This live demo will highlight the journey of an ARC (Annotated Research Context) dataset, coming from the FAIRagro middleware and being consumed by user-facing FAIRagro tools. We will see how researchers can discover it in the FAIRagro Search Hub, and then use SciWIn to amend the ARC with a workflow for execution and further reusable sharing.

More about Julian Schneider & Antonia Leidel
Contributors: Tobias Wamhof, Timo Mühlhaus, Kevin Schneider, Harald von Waldow

The Annotated Research Context (ARC) is DataPLANT’s FAIR Digital Object (FDO) framework that streamlines packaging, annotating, validating, and publishing research data in line with FAIR principles. The ARC framework provides reusable community-tailored metadata templates, automated validation via the PlantDataHub to enable DOI registration and data publication. This quick tour demonstrates the end-to-end workflow: creating an ARC based on templates, performing validation to ensure metadata quality and interoperability, and publishing the resulting FDO via PlantDataHub for dataset visibility and reuse.

The Standards Inventory brings together all important agriculture-related standards, based on the FAIRagro use cases, into a single ontology, complete with a user interface for easy searching and exploration.

More about Nils Reinosch
Contributors: Daniel Martini, Jasha Jung

Agricultural long-term experiments (LTEs) provide valuable multi decadal data for crop modeling and global change research, yet their broader reuse remains limited by the narrow focus of many studies and incomplete adherence to FAIR principles. The LTE Overview Map (https://bonares.de/ltfe/) has compiled metadata for about 680 LTEs worldwide, supporting the reuse of both published and unpublished data and facilitating synthesis studies that combine multiple LTEs. However, the metadata are still incomplete and require substantial manual curation.

We developed an automated pipeline that combines large language models (LLMs) with rigorously defined Pydantic data schemas that guide and control the AI in extracting structured information directly from scientific literature. Implemented as a Python workflow, the system parses papers associated with each LTE, extracts relevant experimental descriptors, and translates them into standardized, interoperable metadata records. In addition, it identifies the research context and specific objectives of each publication using LTE data – information that provides initial indicators for data fitness for purpose and helps track LTE data reuse.

Quality control begins with structural validation, where Pydantic ensures that all extracted fields follow the required format. This is followed by semantic checks that test whether the information makes logical and domain specific sense. Finally, we compare results of 3-5 LLMs and use an AI as judge process that rates the accuracy and reliability of each extraction, supporting transparent and trustworthy metadata generation.

This AI supported approach greatly accelerates metadata curation, improves data discoverability, and supports a more systematic understanding of how LTE data are used across the agricultural research community.

More about Susanne Lachmuth
Contributors: Cenk Dönmez, Carsten Hoffmann, Masahiro Ryo, Gunnar Lischeid

Crop simulation models are inherently complex, requiring multiple steps and integrating numerous variables. This presentation demonstrates the development of standardised, modular crop modelling simulation workflows using the FAIRagro Scientific Workflow Infrastructure (SciWIn) tool. The SciWIn tool enables the creation of portable, reproducible, and shareable workflows.

Contributors: Benjamin Leroy, Kaushik Muduchuru

The FAIRagro Data Quality (DQ) Tool enables researchers in the agroecosystems community to assess the quality of in-field collected data. It is designed for both in-field use, supporting researchers already during data collection, and for post-hoc evaluations of complete datasets to facilitate reuse. Using a brief case study based on time-series data from field-based plant phenotyping (UC5), we showcase the tool’s functionality and practical applications.

More about Sven Gedicke
Contributors: Sven Gedicke, Lars Grygosch, Jan-Henrik Haunert

The FAIRagro Federated Network brings together metadata from different research data infrastructures (RDIs) into a single, coordinated system. It solves common problems like incompatible technologies, varying ways of organizing metadata, and systems that don’t communicate well with each other. A key contribution is the SQL-to-ARC method, which creates standardized Annotated Research Contexts (ARCs) and integrates them across the network.

More about Jorge García Brizuela
Contributors: TA4

The integration of proximal sensing, robotics, and remote sensing into plant phenotyping has created an interdisciplinary data landscape that is vast and complex. While enabling non-destructive measurements under field conditions, these technologies generate heterogeneous datasets that challenge effective data management and interoperability. Current metadata standards, such as the Minimum Information About a Plant Phenotyping Experiment (MIAPPE), provide a crucial foundation but lack explicit constructs for describing sensors, platforms, and computational workflows central to modern, technology-driven phenotyping. This hinders the findability, accessibility, interoperability, and reusability (FAIR) of data and limits the potential for advanced data integration and analysis. To bridge this gap, we propose a novel, ontology-based data model that semantically extends MIAPPE by integrating concepts from the Semantic Sensor Network (SSN/SOSA) ontology. This model formalizes the relationship between biological studies, observational procedures, sensor systems, and derived data, creating a unified framework for documenting interdisciplinary field phenotyping experiments. A primary objective of this model is to serve as a canonical schema for converting disparate, tabular phenotyping metadata into a structured, machine-readable knowledge graph. This graph representation is not only inherently FAIR but also provides the foundational structure for future applications, such as the training of graph neural networks (GNNs) to discover contextual patterns across experiments.

More about Lars Grygosch
Contributors: Lars Grygosch

The FAIR Assessment Tool evaluates the FAIRness of datasets by integrating three existing services. The results are aggregated in an RDF graph using the Data Quality Vocabulary (DQV) and can be downloaded in various formats. They can also be stored in a triple store, making them machine-readable and accessible. This approach allows systematic analysis of FAIRagro repositories and other research data infrastructures. The focus is on assessing the FAIRness of the published resources themselves, rather than the underlying raw data. Current work aims to make the results more understandable for users through a large language model, enabling automated feedback on datasets in natural language.

More about Jascha Jung

Various near-surface geophysical methods are used to characterise soil properties on agricultural sites. In addition, such data provide information on archaeological features and soil dynamics. To make data findable, accessible, interoperable, and reusable (FAIR) for this purpose, appropriate metadata standards are required.

The aim of UC14 is therefore to develop a metadata standard for near-surface geophys-ical prospecting. This standard will later be made available as a tool for data acquisition in order to implement the FAIR principles in practical work and enable interchangeabil-ity. The basis for this is an ontology and the metadata description derived from it.

The model developed in Protégé contains method-independent as well as sensor-specific metadata. During creation, the terms were classified hierarchically and re-strictions were introduced to ensure that basic information would be captured later and that there would be scope to add additional information. Furthermore, data and object properties were defined for the terms in order to specify data formats and also to use defined terms from predefined lists. Harmonisation of defined terms with existing standards has also been initiated in order to achieve uniform use of selected parame-ters and thus improve the comparability of the data sets described.

The metadata model presented here represents an initial development stage. It com-prises method-independent metadata as well as specific models for electromagnetic induction (EMI) and magnetic prospection.

The initial released model will be discussed at the community summit in order to identi-fy potential improvements for the future and optimise the model for later application.

More about Johannes Rabiger-Völlmer
Contributors: Johannes Rabiger-Völlmer, Till Sonnemann, Claudia Schütze, Ulrike Werban

This study evaluated the use of keywords to annotate scientific literature metadata. FAO’s agricultural literature database FAO AGRIS is used as an example. FAO AGRIS has more than 16 million bibliographic records of food and agricultural scientific literature in 258 languages. The data is harvested from more than 1000 institutional repositories and publication databases from around the globe. About 80 % of these records are annotated with keywords. As they are provided in the AGRIS dataset as free text strings, the source of these keywords is unknown. A large percentage however can be found in AGROVOC, FAO’s multilingual thesaurus on food and agriculture. The study aimed to find reasons why not all keywords in AGRIS are taken from AGROVOC. Possible reasons are: language of keyword is not covered, keyword is not in the scope of AGROVOC or gaps in the content coverage of AGROVOC. The tool created is not only suited to metadata sets but might also be used in the future within FAIRagro for doing large scale (semi-)automated ontology alignment.
Apart from insights into metadata annotation practices, the study showed how the use of keywords from a controlled vocabulary such as AGROVOC to annotate scientific literature data records eliminates the need for computationally expensive string matching. It provided future directions for filling gaps in AGROVOC (content and languages). AGRIS data providers should be encouraged to use keywords from AGROVOC as well as repository developers and providers should provide facilities to annotate using controlled vocabularies.

Animal phenotyping (e.g., FAANG, PigWeb) generate vast datasets critical for genomics, welfare, and precision livestock farming. However, inconsistent metadata collection hampers data integration, reproducibility, and reuse across species and projects.
We propose a unified minimal metadata standard for animal phenotyping by adapting the MIAPPE (Minimum Information About Plant Phenotyping Experiments) checklist—a well-established framework in plant science—to address animal-specific requirements.

Extensive data on soil, climate (or weather), and the history of agricultural measures are required for modeling and predicting physical, chemical, and biological soil properties in agriculturally used soils. BODIUM4Farmers is a web-based modeling tool that offers efficient methods for integrating existing databases for practical agriculture. This article presents the concepts that have been implemented.

Für die Modellierung und Vorhersage von physikalischen, chemischen und biologischen Bodeneigenschaften in landwirtschaftlich genutzten Böden werden umfangreiche Daten zu Boden, Klima (bzw. Wetter) und zur Historie der landwirtschaftlichen Maßnahmen benötigt. BODIUM4Farmers ist ein webbasiertes Modellinstrument, welches effiziente Methoden zur Einbindung vorhandener Datenbanken für die praktische Landwirtschaft bietet. In diesem Beitrag werden die implementierten Konzepte vorgestellt.

More about Hans-Jörg Vogel

At the Research Institute for Farm Animal Biology (FBN), we have established a Data Service Centre (DSC) to support the full lifecycle of research data, ranging from experimental design and the preparation of Data Management Plans to reproducible statistical analysis, standardized data structures, long-term archiving, and the publication of data and software. A central aim of the DSC is to support researchers in implementing FAIR principles in their daily work, while avoiding unnecessary additional workload. We also address current challenges in agricultural research, including the still limited availability of established metadata standards and domain-specific repositories for livestock data. Our work is closely linked to active participation in the National Research Data Infrastructure (NFDI), where we both benefit from and contribute to community-driven developments and the growing network of expertise fostered by FAIRagro and related initiatives.

Despite their potential benefits of enhancing environmental and economic performance, it remains unclear which diversified systems perform best and how feasible their adoption is for farmers. PhenoRob2 addresses these challenges by leveraging technologies to design diversified farming systems. Addressing the most beneficial target system requires an integrated assessment of agronomic outcomes, as well as adoption constraints and technologies required to enable their successful implementation. However, there is surprisingly limited empirical evidence on the economics of diversification. The overarching objective of my PhD therefore is to provide robust theoretical foundations and empirical evidence on their economics and farm-level drivers of adoption.
 
My PhD project is closely embedded within CP5 and will interact with other doctoral projects to jointly define key economic and adoption indicators relevant for the possibility space of diversified systems. Further, exchange with agrosystem researchers and RDM experts will be essential for systematically collecting existing data and for identifying knowledge gaps.
 
First, large existing datasets will be synthesized to derive generalizable insights into productivity, costs, risk, and adoption patterns. Second, CP5 field-level data will be combined with farmer surveys and economic experiments to capture behavioral dimensions of adoption.
 
The project is expected to deliver (i) key economic and adoption indicators for CP5, (ii) a theory on the economies of scale and scope in diversified systems, (iii) an empirical assessment of economic performance and adoption drivers, (iv) these will be key foundational inputs for developing integrated bio-economic models for out-of-sample predictions of economic outcomes and adoption potential of diversified systems.

Scientific data volumes continue to increase, requiring clear decision processes to identify data for deletion and data for long-term preservation beyond the ten-year minimum mandated by good research practice. CKAN was assessed as an institutional, open-source system for managing and archiving internal research datasets. The evaluation focused on design approaches and extensions that support controlled archiving workflows and facilitate subsequent dataset publication, including export to platforms such as Zenodo. The poster presents these concepts and demonstrates how CKAN can be integrated into institutional research data management to support sustainable data preservation.

Contributors: Steffen Franke, Yana Bodnar, Rainer Cramm, Christopher Rickelt, Wolgang Clauss

Between the acquisition of raw data and its publication lies considerable work and time. Data must be checked, filtered, calibrated, and analyzed. Experience from various collaborative projects and consulting practice as a research data steward at Justus Liebig University shows that this work is often poorly documented, and data is stored in an unstructured manner. For many measurements in the agricultural and environmental sectors, contextual information is also crucial for accurate evaluation: measurement instruments require maintenance, study areas are cultivated or devastated by storms.
To preserve such information over the long term, the applicants have developed the open-source field data management software Observatory Data-Management Framework (ODMF). ODMF is a web application with a data model for storing time series data at geographically locatable points. Data can be transparently maintained: erroneous measurements are marked rather than deleted; calibrations are calculated on-the-fly; derived quantities are computed directly. The web application is used for visualization and data export. Sensor networks with telemetry, such as LoRa, ADCON, and Campbell systems, can serve as data sources. Other data sources, like the Meteosol climate station network, are integrated via the ODMF API. Output formats from offline loggers, laboratory analysis devices, or legacy templates for field management can be described with configuration files and imported into the database annotated with metadata.

Climate and environmental data come from many sources and are often difficult to work with due to inconsistent formats and structures. climdata is a lightweight, open-source package that helps harmonize and process these datasets in a consistent, analysis-ready way. climdata promotes clear, standardized workflows that make tasks such as extreme index calculation, data imputation, bias correction, and climate model selection easy to execute through chained workflows.

Contributors: Kaushik Muduchuru, Amit Srivastava, Frank Ewert (Multi-Scale Modelling & Forecasting, ZALF)