Environment, Natural Resources and Agriculture research strategy 2027-2032: consultation analysis

10. Question 8

Theme: Impact Routes – Decision Support, Living Labs and innovation

Question 8: Which actions relating to data, data analysis, and modelling should the ENRA research programme prioritise?

Introduction

The majority (47, 67%) of consultation respondents answered question 8.

The vast majority of respondents agreed with the desired modelling and underpinning data analysis outcomes

The vast majority of respondents (all respondent types) who answered question 8 said the research programme should prioritise actions that were aligned to or in broad agreement with the outcomes outlined in Annex C of the consultation document. The three outcomes are:

• ensuring traceability and transparency in how data, analysis, and models inform decisions
• enabling timely, evidence-based policy decisions through accessible, high-quality analysis and modelling
• fostering a culture of reuse, collaboration and active engagement across research and policy communities

Theme 1: Ensuring traceability and transparency in how data, analysis, and models inform decisions

The majority of respondents (mostly organisation respondents and in particular Research Institutes and Centres of Expertise and other scientific organisations) said that improving traceability and transparency of data, data systems and modelling represented a clear and credible step forward for the programme.

Specific actions that respondents felt the ENRA research programme should prioritise included:

• providing outputs in accessible formats, such as scenario tools, disaggregated data and policy summaries – respondents said this could make the research programme more visible and easier to communicate to range of different stakeholders
• open publication of workflows
• improved data and documentation
• improved/clearer data provenance
• validation against real world data
• a coherent governance framework for data modelling, covering uncertainty, scope and limitations to ensure consistent metadata standards – respondents said this could make data analysis easier to understand, scrutinise (for example, peer review) and reuse

Further, respondents noted that processes to support transparency and traceability could bring greater clarity to how evidence is gathered, generated and interpreted. In turn, this could help ensure greater alignment with policy, embed long-term monitoring and evaluation, avoid duplication and improve trust and confidence in decision-making. These organisation respondents viewed building in traceability and transparency processes as a practical and well-established approach across national and international data systems.

“Develop models that are transparent, easy to update, and can be adapted to different scenarios, scales, and user groups, including rural planners and local decision-makers.” National Sheep Association

“Combining existing component models is a good start to the development of systems models, but they can limit the range of applicability due to computational bottlenecks, statistical difficulties in parameter estimation, and missing feedback mechanisms that may lead to a failure to identify unintended outcomes or win-wins. Therefore, robust decision support also requires development of simplified and computationally tractable systems models that can be subjected to more rigorous statistical testing and analysis, and such efforts should be supported by the ENRA programme.” Biomathematics and Statistics Scotland

Theme 2: Enabling timely, evidence-based policy decisions through accessible, high quality analysis and modelling

Many organisation respondents (particularly Research Institutes and Centres of Expertise) stated that the research programme should prioritise accessible high quality data and analysis in order to ensure timely and targeted evidence for decision makers.

These respondents regarded the accessibility of high quality data as an effective method to ensure that the research programme is closely aligned to the Missions and ARIs and is adaptable to the pace of environmental and societal change. For example, Living Labs and regional testbeds were consistently highlighted by organisations (particularly Research Institutes) as valuable platforms for embedding engagement and creating iterative learning and feedback loops to help inform both policy and practice.

A related point raised (all respondent types, in particular other public bodies) was that data curation was essential to identify gaps and opportunities for integration and collaboration within data infrastructure. Data curation was also considered to be an essential quality control method to ensure more responsive and robust decision making such that there is ‘the pathway from data to decision making.’

Some of these respondents called for greater investment in technical infrastructure for advanced modelling, such as AI and Light Detection and Ranging (LiDAR), to enable timely analysis of data that is robust, relevant and available when needed. This included developing models that are adaptable to different scales and scenarios (for example, from farm, plot and catchment scale to Scotland level).

“Considerable gains could be made by more emphasis on high quality modelling and use of existing data. Possibly looking at models developed elsewhere which, if necessary, could be modified for Scottish conditions. The Scottish Research Institutes have excellent long term data sets and collections that are of great value in developing evidence-based models.” Hannah Dairy Research Foundation

“We would strongly advocate the exploration of AI in addressing the food system challenges outlined in this strategy, building on work that has already been delivered through the SRP and research commissioned by Food Standards Scotland and other funders. This includes the application of bioinformatic machine learning approaches for understanding pathogen transmission and the development of tools that improve our ability to detect changes in the food landscape (such as changes to production methods, food composition and consumption patterns) which support the evaluation of policy (for example, indicators for assessing performance against the goals of the Good Food Nation Plan).” Food Standards Scotland

“The opportunities to utilise AI should form a strong focus within the strategy. FAIR (Findable, Accessible, Interoperable, and Reusable) data standards, shared infrastructure and investment in modelling and decision-support tools are vital to unlock innovation across the supply chain.” The Maltsters Association of Great Britain (MAGB)

Theme 3: Fostering a culture of reuse, collaboration and active engagement across research and policy communities

Many organisation respondents (all organisation types) highlighted that active engagement and collaboration between researchers, policymakers, practitioners, and communities is fundamental to achieve the Vision for Modelling in the SRP.

They suggested that priority actions could focus on training and capacity building (such as mentoring), particularly for community groups and farms, to support communities and the wider general public to be involved in data, data analysis and modelling. It was noted that this could ensure greater adoption of the research outputs.

Other suggested priority actions included enabling co-designed tools, integration of citizen science platforms, shared workflows and multi-disciplinary teams. Respondents said that these could support systems thinking, highlight opportunities for partnership and collaboration and allow lessons to be learned and shared across different Missions and Challenges.

Some respondents placed value on actions and approaches which would make best use of existing high quality datasets, such as common data standards, integrated and interoperable models and coordinated access to existing datasets. They noted that this could provide benefits such as efficient use of public resources by avoiding duplication. Respondents said this would build capacity and foster collaboration to deliver a more connected, coherent and impactful research programme building on the extensive work already achieved by the Research Institutes.

Further, some of these respondents pointed to the value of linking Scottish datasets and analytical aproaches with UK and international initiatives to strengthen comparability and open further routes for collaboration. They noted that there are many areas where data requires an international approach, such as pathogen genomics, food system monitoring, climate scenario standardisation and crop genetic resources.

“The 2027-2032 Strategy builds upon expertise and stakeholder networks developed within the MRPs over many years. Such expertise and shared understanding of policy and research needs will enable knowledge to be advanced while adding value to existing capabilities and infrastructure (e.g. research platforms, specialist facilities). Where partnerships with other researchers or research organisations are required, these are either already in place and operating well or can be identified and involved via the MRPs.” James Hutton Institute

“Data will deliver far greater value when integrated and shared across national boundaries. The ENRA research programme should therefore prioritise collaboration on data and modelling with partners across the wider GB/UK and internationally. This is particularly important in areas such as animal and plant genetics, and in carbon auditing to support transparent supply chain reporting. A “collect once, use many times” principle should underpin data collection, ensuring that permissions and data governance are appropriately managed - especially where farmers are asked to contribute. There is a risk that excessive data demands could create unnecessary burdens on farmers and other industry partners, potentially discouraging participation.” Agriculture and Horticulture Development Board