REHUBS

Project Description

REHUBS addresses the critical issue of climate vulnerability in rural areas and small-to-medium-sized communities, which are particularly susceptible to climate risks such as droughts, heatwaves, wildfires, floods, and biodiversity loss. The project’s goal is to enhance the climate resilience of these areas by developing an integrated framework of adaptation solutions, advanced digital tools, and participatory approaches, specifically tailored to the rural context and landscape and designed to be easily implemented and practiced by non-experts.

The project designs, implements, and validates a comprehensive suite of tools including a climate risk assessment model, an advanced early warning system, flood and drought modelling tools, smart irrigation systems, a circular water recovery and treatment system, a rapid AI-based wildfire detection system, and a Risk Assessment and Adaptation Platform (RAAP) that recommends low-cost Nature-Based Solutions (NBS) for rural resilience. REHUBS also promotes knowledge transfer and upskills local actors through an online academy, an AI support chatbot, and collaborative Rural Micro Hubs (RMH) that bring together local authorities, communities, and stakeholders.

The solutions are implemented and validated across four demonstration clusters, with replication across three additional sites, integrating climate adaptation strategies into local governance frameworks. The project also introduces innovative financial instruments to ensure the cost-effectiveness and long-term replicability of its solutions across diverse European regions.

LINKS Foundation Activities

LINKS Foundation participates as a research and technology partner, contributing expertise in Artificial Intelligence, Earth Observation data analysis, and machine learning applied to wildfire detection, climate risk forecasting, and rural resilience tools.

AI-Based Rapid Wildfire Detection System

Design and development of an AI-powered wildfire detection system leveraging advanced machine learning techniques, including Self-Supervised Learning (SSL), to accurately detect and classify wildfire hotspots in near real-time from Earth Observation data (Copernicus, VIIRS, MODIS), enabling:

• automated early-stage wildfire identification to support prompt response and fire suppression
• deep learning models processing time-series and multimodal satellite data for accurate hotspot detection
• data fusion techniques to enhance real-time detection capabilities and reduce false alarms

Integrated European Wildfire Risk Forecast

Development of an AI-driven dynamic forest fire risk index generating daily risk forecast maps with up to 10 days of lead time, trained on data from the Copernicus programme, ECMWF, Eurostat, and a proprietary LINKS land cover model, supporting:

• predictive fire ignition modelling and fire behaviour simulation using machine learning and deep learning approaches
• delivery of risk maps to local municipalities for real-time wildfire planning and decision-making
• integration of outputs into the CONFIRM real-time wildfire information sharing platform developed by partner VOST

Risk Assessment and Adaptation Platform (RAAP)

Contribution to the architecture and development of the RAAP platform, which integrates climate risk data, AI tools, and adaptation recommendations into a single interface for rural authorities and local actors, supporting:

• definition of functional and non-functional platform requirements (performance, scalability, security, privacy)
• integration of AI chatbot, climate regional projections, advanced weather modelling, NBS framework, online academy, and wildfire detection outputs into a coherent system
• delivery of a user-friendly, accessible tool enabling non-expert users to assess climate risks and identify appropriate low-cost adaptation measures

Climate-Related Early Warning System

Support to the development of an advanced integrated early warning system for rural resilience against multiple climate hazards — droughts, floods, heatwaves, and forest fires — contributing to alert configuration, risk model integration, and local preventive protocol development.

Funding & Dates

SDGs — Sustainable Development Goals

🏙️ Goal 11 — Sustainable Cities and Communities

🏗️ Goal 9 — Industry, Innovation and Infrastructure

🌍 Goal 13 — Climate Action

Key Technologies

🤖 AI & Deep Learning for Wildfire Detection

Self-Supervised Learning (SSL), deep learning, and data fusion techniques applied to multi-source Earth Observation data (Copernicus, VIIRS, MODIS) for near real-time wildfire hotspot detection and classification, delivering automated early-stage alerts to emergency management authorities.

🛰️ Earth Observation & Copernicus Data

Exploitation of Copernicus satellite data streams alongside ECMWF meteorological data, Eurostat datasets, and proprietary land cover models to train and validate AI-based wildfire risk forecasting models providing daily risk maps with up to 10 days of lead time.

🌿 Nature-Based Solutions (NBS) Framework

Development of a structured, evidence-based catalogue of low-cost Nature-Based Solutions for rural climate resilience, integrated into the RAAP platform to recommend practical adaptation measures tailored to local conditions and accessible to non-expert users.

🖥️ Risk Assessment and Adaptation Platform (RAAP)

An integrated digital platform combining AI chatbot support, climate risk projections, advanced weather modelling, wildfire detection outputs, and an online training academy into a single user-friendly interface designed for rural local authorities and community stakeholders.

🚨 Integrated Early Warning System

A multi-hazard early warning system covering droughts, floods, heatwaves, and forest fires, combining real-time sensor data, risk model outputs, and alert dissemination tools to enhance community preparedness and emergency response in rural areas.

👥 Rural Micro Hubs & Participatory Approaches

Structured local collaboration nodes bringing together municipalities, farmers, civil society, and volunteers to co-design, test, and validate REHUBS solutions, ensuring that tools are practically usable by non-experts and embedded in local governance processes.

Project Partners