Projects
LIGHTEN-IO
Our century is marked by a race against time to limit global warming to 1.5 °C above pre-industrial levels, as agreed under the Paris Agreement by 192 Parties in December 2015. In this context, the structuring of regional green hydrogen value chains is a major strategic challenge, particularly for the island territories of the South-West Indian Ocean (SWIO), whose abundant yet intermittent renewable resources hold considerable — and largely untapped — potential.
Accurately assessing this potential, however, runs up against a central bottleneck: the effective accessibility of the climate and energy data required to characterise renewable resources. Where such data exist, they are scattered across institutions, produced according to heterogeneous standards and stored in disparate formats, which hinders their discovery, interpretation and reuse. The island territories concerned also suffer from observation gaps and incomplete time series, limiting the reliability of analyses. To these challenges are added the scientific hurdles of climate downscaling: computational constraints, stochasticity of deep learning models, the capacity to generalise beyond the training period, and the need to preserve the physical consistency of high-resolution reconstructed fields.
To overcome these obstacles, an approach combining data engineering, climate modelling and machine learning is now being mobilised. U-Net neural network architectures, coupled with ingestion, normalisation and interoperable distribution pipelines (OPeNDAP, THREDDS), make it possible to transform a fragmented set of resources into a coherent, well-documented and easily exploitable ecosystem. These approaches pave the way for a reliable assessment of green hydrogen production potential and for the development of decarbonisation scenarios tailored to territorial realities.
The LIGHTEN-IO project — Leadership Initiative for Green Hydrogen Transition Energy Network in the Indian Ocean —, scientifically led by Prof. Béatrice Morel, is part of this dynamic. It aims to structure a regional network of stakeholders around green hydrogen in the SWIO area, to generate reference regionalised climate and energy data, to assess green hydrogen production potential across the target territories, and to build sectoral decarbonisation scenarios. The project brings together an international consortium including the Université des Mascareignes, the University of Mauritius, IST-T Antananarivo, the University of Nairobi, the Université des Comores, Forschungszentrum Jülich, Météo-France and the Seychelles Meteorological Authority.
The LIGHTEN-IO project is funded with a total budget of €598,939.78 under the INTERREG VI Indian Ocean 2021-2027 programme (Action Sheet 1.3), with a €509,098.80 (85%) contribution from the ERDF and an €89,840.98 (15%) matching contribution from the Région Réunion. Europe is committed in Réunion through ERDF funding.
This project strengthens local skills at the interface of climate sciences, energy and data sciences, and contributes to the structuring of a regional green hydrogen value chain serving the energy transition and the resilience of island territories in the South-West Indian Ocean.
The LIGHTEN-IO project builds on the achievements of the SWIO-Energy partnership (2020-2023) and the IOS-net network (2019-2022) to move from bilateral scientific exchanges to the structuring of a genuine regional network dedicated to green hydrogen. It is organised around four specific objectives:
1. Structuring and coordinating a SWIO stakeholder network
The objective is to design a concerted strategy for the deployment of green hydrogen across the South-West Indian Ocean. Several tasks are carried out:
- Organisation of a kick-off seminar in Réunion and definition of deliverables, roles and milestones;
- Development of a shared regional strategy among SWIO partners;
- Establishment of a monitoring system for territorial initiatives;
- Coordination of regular training workshops and working meetings with partners.
2. Generation of regionalised climate and energy data
The objective is to produce harmonised, open datasets enabling the fine-grained assessment of renewable resources. Several tasks are carried out:
- Collection, ingestion and normalisation of climate, meteorological and energy data from in-situ observations, satellite products and climate simulations;
- Training in downscaling tools (LSCE) and development of a statistical downscaling tool based on U-Net architectures (Keras/TensorFlow);
- Production of climate fields at high spatial and temporal resolution from global model outputs;
- Distribution of data via an open-access THREDDS Data Server, accompanied by a data management plan and a data paper.
3. Assessment of green hydrogen potential and decarbonisation scenarios
The objective is to leverage regionalised data to support specific case studies and inform public decision-making. Several tasks are carried out:
- Training in the RESkit tool (IEK-3, Jülich) and assessment of green hydrogen development potential;
- Study of climate change impacts and land-use changes;
- Comparative analysis of SWIO territories and development of sectoral decarbonisation scenarios;
- Organisation of scientific conferences on the regionalisation of renewable energies and green hydrogen production in the SWIO area.
4. Dissemination, outreach and regional structuring
The objective is to ensure the dissemination of results and the sustainability of the network. Tasks include:
- Publication of scientific articles (statistical downscaling, green hydrogen potential, decarbonisation scenarios);
- Open-source release of source code and datasets;
- Production of a project website and a documentary series for outreach purposes;
- Organisation of the final seminar and training workshops in target territories.
The LIGHTEN-IO project aims to deliver concrete, reusable outputs with strong scientific and territorial impact across the SWIO region.
The main expected outcomes are:
- An open, interoperable database: collection, harmonisation and distribution of a significant volume of climate, meteorological and energy data through a THREDDS Data Server, accompanied by enriched metadata and a data management plan. This resource will provide a valuable asset for the scientific community and energy-transition stakeholders across the South-West Indian Ocean.
- A deep-learning-based statistical downscaling tool: development of a pipeline using U-Net architectures, enabling the generation of regionalised climate projections at high spatial resolution with lower computational cost compared to traditional dynamical downscaling approaches. The source code will be released as open source to foster reproducibility and replicability to other island regions facing similar challenges.
- Territorialised energy and decarbonisation scenarios: assessment of green hydrogen production potential across SWIO territories, analysis of climate change impacts on renewable resources, and development of sectoral decarbonisation scenarios designed to inform public policy.
- Scientific advances and regional outreach: publication of scientific articles (statistical downscaling, green hydrogen potential, decarbonisation scenarios), organisation of two thematic conferences, and delivery of a final seminar. These outputs will strengthen ENERGY-Lab’s positioning as a regional hub on green hydrogen issues in insular environments.
- Capacity building and regional structuring: cross-training between partners (LSCE, IEK-3/Jülich), workshops in target territories, and the establishment of shared regional governance. In doing so, the project transforms a still-blurred regional vision into a precise, operational local strategy for the deployment of the green hydrogen value chain.
Through these deliverables, LIGHTEN-IO will contribute to reducing the cost of access to strategic climate information, assessing the viability of green hydrogen as a regional energy carrier, and enhancing the energy resilience of the island territories of the South-West Indian Ocean.
Financial partners
The LIGHTEN-IO project is funded by the European Union under the INTERREG VI Indian Ocean 2021-2027 programme (Action Sheet 1.3), with the Région Réunion acting as Managing Authority. Europe is committed in Réunion through ERDF funding, complemented by a matching contribution from the Région Réunion.
Academic partners
- Université des Mascareignes (UdM) — Mauritius: expertise in renewable energies (solar, wind), energy efficiency and dynamical downscaling. Previous SWIO-Energy partner (2020-2023).
- University of Mauritius (UoM) — Mauritius: expertise in renewable energies and both dynamical and statistical downscaling. Previous SWIO-Energy partner (2020-2023).
- Institut Supérieur de Technologie d’Antananarivo (IST-T) — Madagascar: expertise in renewable energy systems and energy transition in Madagascar; gateway to local energy professionals.
- University of Nairobi (UoN) — Kenya: expertise in hydrogen production pathways and artificial-intelligence methods applied to statistical downscaling.
- Université des Comores (UC) — Comoros: gateway to academic and professional energy stakeholders in the Comoros. Previous IOS-net partner (2019-2022).
Institutional partners
- Forschungszentrum Jülich (IEK-3) — Germany: expertise in renewable potential assessment tools (GLAES, RESkit) and associated training.
- Météo-France — Réunion Island: access to regional meteorological data.
- Seychelles Meteorological Authority (SMA) — Seychelles: provision of SMA station data and expertise on South-West Indian Ocean climate. Previous IOS-net partner (2019-2022).
- Scientific coordinator:
- ENERGY-Lab management team involved in the project: