ARIA: Forecasting Tipping Points Phase 2 – Concept papers

ARIA welcome applications from across the R&D ecosystem, including individuals, universities, research institutions, small, medium and large companies, charities and public sector research organisations.

For the vast majority of applicants, we require the majority of the project work to be conducted in the UK (i.e. >50% of project costs and personnel time).

We can award funding to applicants whose projects will primarily take place outside of the UK if we believe it can boost the net impact of a programme.

Join a team

We have a live teaming tool that allows applicants to find complementary expertise. After a quick registration, you can browse other researchers and request an introduction from the ARIA team to explore potential collaborations.

Phase 2 aims to establish whether a trustworthy, decision-relevant prototype EWS can be built for the SPG, with freshwater forcing from Greenland ice sheet meltwater as a key input. The Forecasting Tipping Points objective is not to guarantee prediction of a tipping threshold being crossed, but to show whether the evolving Earth system can be sufficiently well characterised to provide useful evidence about changing risks and decision windows. We therefore test explicitly whether the relevant SPG state can be sufficiently constrained, whether warning and forecasting methods are robust, and whether the resulting information can support a credible response, through three repeated learning cycles that aim to achieve the following programme-wide objectives:

1. Demonstrate and improve an observing-to-decision chain.
2. Improve evidence on SPG state and freshwater forcing.
3. Deliver trustworthy warning and foresight.
4. Use evidence to improve observing and response decisions.
5. Leave an adoptable service.

No individual project is expected to deliver the entire EWS (early warning system): successful projects will contribute to a shared, interoperable system. We welcome proposals spanning more than one module where this is necessary to close an interface or deliver a coherent product.

Every proposal should identify its relevant module(s); its expected inputs and outputs; its named downstream consumer; the acceptance test for its outputs to be verified; and how uncertainty, validation and data provenance will be managed.

What we expect to fund

We want to see ambitious proposals in four particularly important areas (with indicative budgets for the whole module): for full details of each area, read the call for proposals.

An ocean observing network (£18.5m): Priority gaps include observing freshwater distribution, ocean-atmosphere exchange, stratification, deep convection, circulation and mixing; deployment of new moorings with timely readout; and approaches that use observing-system experiments, state estimation and forecasts to target measurements where they reduce the most important uncertainties.

Interannual to decadal forecasting (£5-7m): a system that is initialised and constrained by the evolving state of the SPG, produces useful information on interannual-to-decadal timescales, and can be tested and improved through repeated use. We are seeking physics-based and AI-native approaches that connect directly to the programme’s state-estimation outputs; make uncertainty, assumptions and failure modes explicit; and produce physically consistent forecasts or storylines of timing, magnitude and geography of changes. Proposals should include a credible evaluation strategy, including comparison with relevant external baselines (e.g. the Decadal Climate Predictions Project models) and matched no-new-data analyses where appropriate.

Impacts and response (£3-5m): We are seeking teams with embedded user communities, sector expertise and product-development capability to co-design, test and refine decision-facing outputs.

Independent systems integration and assurance (£5.8m): proposals for independent systems integration to create and maintain the shared EWS MVP (early warning system minimum viable product). Systems integration will connect data, models, analytical tools and teams across the modular architecture; manage module interfaces and dependencies; establish integration cycles, acceptance tests, provenance and release processes; and ensure that uncertainty and fallbacks remain visible through the system.

Other modules and portfolio fit: Meltwater observation, state estimation and tipping point signal detection remain essential elements of the programme and will be integrated into the Phase 2 portfolio through existing programme capability. While this solicitation is not intended to duplicate existing capability in these areas, new proposals may enter these modules where they address a clearly evidenced, decision-critical gap and have a defined upstream or downstream role.

We are unlikely to support:

• isolated sensors, short field campaigns or observational datasets without a credible route into state estimation or forecasting
• standalone modelling or AI approaches that do not connect to observations, end-user requirements and prospective validation
• generic dashboards, data platforms or communication products without a defined decision, user workflow and route to use
• stakeholder engagement that does not include co-design, testing or progress towards operational adoption
• proposals that seek to work in Greenland with insufficient Greenland engagement
• competing end-to-end EWSs developed independently of the programme’s shared interfaces and integration process