Prototype database & exploration infrastructure

Integrated agri-food policy analysis requires the combination of data and evidence produced by different sources, models and analytical modules. Without common structures and traceable workflows, data processing can become difficult to reproduce, outputs can be difficult to compare, and connections between different stages of analysis can become harder to maintain.

ACT4CAP27 addresses these challenges by providing a shared environment in which data, processing routines, model components and outputs can be managed through common technical procedures. The infrastructure is designed to support modularity, reproducibility, interoperability and scalability as new datasets, models and analytical components are added.

ncluding approaches for version-controlled data management, modular processing, automated workflows, defined software environments and standardised model outputs.

D4.2 documents the transition from this technical foundation to a functional prototype on the IIASA Accelerator platform. The prototype brings together a Data Repository and a Scenario Explorer.

The Data Repository provides a central environment for storing and managing datasets and model outputs. Git and Data Version Control (DVC) support versioning and traceability, while automated validation routines check datasets against defined reporting templates. The infrastructure can also preview and explore common data formats and supports workflows for publishing datasets to Zenodo.

A flexible reporting template based initially on the AgMIP standard provides common structures for variables, units and regional definitions. The template is being extended towards an ACT4CAP27-specific format to support both model-output comparison and data-input harmonisation.

The Scenario Explorer provides a web-based interface for browsing and analysing datasets and scenario results. It supports interactive visualisation of regional and spatial data, including time series, maps, Sankey diagrams and other chart types, with filtering by indicators, regions and time periods.

The prototype also includes AI-assisted data exploration, allowing users to query datasets using natural language and receive summaries of patterns and trends. This function supports exploration and interpretation of data while remaining part of a broader research workflow in which datasets, processing and analytical assumptions can be inspected and verified.

Together, these components establish an input-to-output infrastructure linking data management, processing, model results and interactive exploration.

The infrastructure provides a common technical basis for connecting data and modelling work across ACT4CAP27. Version control and traceability help users identify the data and processing associated with a particular analysis, while common templates and validation routines support more consistent handling of outputs from different modelling teams.

The modular architecture also allows individual datasets, processing routines, model stages and analytical components to be developed or updated independently. This supports the project’s wider work on connecting complementary modelling approaches and indicators without requiring all analytical components to operate within a single model or software environment.

The Scenario Explorer adds an accessible layer for examining complex datasets and results, helping researchers move from stored model outputs towards interactive comparison and analysis.

The primary users are ACT4CAP27 partners involved in data preparation, modelling, model integration and analysis. This includes researchers, modelling teams and data specialists working with economic, environmental, climate, socio-economic and health-related indicators.

The infrastructure also provides a basis for wider use of ACT4CAP27 data and results as the project develops its public data catalogue and final stakeholder-facing Portal.

The infrastructure can support a range of activities across the ACT4CAP27 workflow, including:

• storing and versioning raw, processed and model-generated data;
• transforming raw datasets into model-ready inputs;
• validating datasets against common templates;
• connecting outputs from different models and analytical modules;
• exploring indicators and scenario results by region, period or variable;
• comparing data and results using common structures and visualisations;
• creating and sharing collections of charts and maps for collaborative analysis;
• supporting technical data auditing through integrated development environments and notebooks;
• preparing datasets for publication and citation through Zenodo.

The Scenario Explorer can also support exploratory analysis of policy scenarios by making complex datasets and model outputs accessible through interactive visualisation and natural-language data queries.

The infrastructure is being developed as part of the ACT4CAP27 project and integrated into subsequent modelling and analytical activities. Technical guidance and use cases established through D4.1 support adoption by consortium partners, while the prototype documented in D4.2 demonstrates the operational environment.

Public access will be provided progressively through the ACT4CAP27 website and Zenodo resources. D4.2 was submitted to the European Commission in April 2026 and will be published through the project’s Zenodo community once it has received EC approval. Further development of the computational framework, dataset catalogue and stakeholder-facing Portal will extend access and usability during the remaining project period.

The technical foundations and a functional prototype are established. D4.1 provides the technical guidance for modular data processing, version control, reproducibility and interoperability, while D4.2 demonstrates these principles through an operational prototype combining the Data Repository and Scenario Explorer.

The infrastructure is therefore already being used within the project environment for data management, processing and exploration. At the same time, it should be understood as a prototype within an ongoing development pathway. The final ACT4CAP27 computational framework and stakeholder-facing Portal will build on this foundation, extending automation, model integration, data access and user-facing functionality.

The infrastructure can support improvements in data management, modelling workflows, reproducibility, model interoperability and scenario analysis within integrated agri-food policy research.

By providing common technical structures for data and model outputs, it can also facilitate collaboration between modelling teams and support the reuse, comparison and further analysis of project results. Its potential contribution to policy analysis depends on the quality, scope and interpretation of the underlying datasets and modelling results.

Stakeholders can engage with the infrastructure through the data, scenario results and analytical outputs made available through ACT4CAP27. As the Portal and its public-facing components develop, these resources can provide opportunities to explore indicators, compare scenario results and examine evidence relevant to future agricultural and food-system policy questions.

The infrastructure also provides a basis for making analytical workflows and results more transparent and reusable. Stakeholder engagement can therefore complement the technical infrastructure by helping identify relevant policy questions, indicators and analytical needs for future applications.