Innovations
Where Earth Observation becomes Intelligence
e-genesis brings together 16 innovations for Europe across biodiversity, water, urban and energy resilience, combining Earth Observation, AI and advanced modelling to push Earth Intelligence beyond the state of the art.
e-genesis Innovation Map
Cross Cutting Core Innovations
1
Co-design
2
Agentic AI
Biodiversity
3
Ecosystems
4
Ecosystem Health
5
Woody Elements
Water
6
Water
7
GenAI Hydro
8
Water Demand
9
Streamflow
10
Snow
11
Groundwater
Urban & Energy
12
Energy Drought
13
Dust
14
Heat
15
PM (Particulate Matter)
16
CO₂
Innovations
Explore the Innovations
Co-design for Earth Observation
What is it?
A collaborative methodology for designing Earth Observation (EO) solutions applied by the consortium together with scientists, service developers, stakeholders and users.
What challenge does it address?
There is a substantial distance between EO specialists and users causing disruption in the process of the commercial uptake of services and their consequent societal impact.
What’s the breakthrough?
Building on the work done in the H2020 project e-shape, the subsequent DestinE programme and its methodology kit, e-genesis aims to move beyond co-designing individual services to co-designing integrated Earth Intelligence solutions, involving a diverse, multilateral community of users and stakeholders.
Why does it matter?
Following a co-design methodology applied horizontally in all three pillars, ensures that the solutions are user-driven, easily transferable and ready for practical implementation in the real world.
Agentic Earth Intelligence for Decision Support
What is it?
An AI-powered decision-support framework incorporating Earth Observation (EO) data, analyses, workflows as well as scientific and policy knowledge.
What challenge does it address?
EO analysis can be complex, requiring users to find the right data, tools and expertise before meaningful insights can be produced.
What’s the breakthrough?
Leaving behind prompt-based workflows, GenAI paired with Earth Observation materialises Earth Intelligence, moving towards agentic AI, GeoGraphRAG and knowledge grounding. The result is an end-to-end decision-support framework, with improved interpretability, transparency and trust, that can discover data, retrieve evidence and generate traceable recommendations.
Why does it matter?
Transitioning from Earth Observation to Earth Intelligence, e-genesis brings about new levels of understandable, evidence-based and intelligence applicable to the project’s three resilience pillars.
AI Seasonal Prediction of Ecosystem Responses to Climate Variability
What is it?
An AI seasonal prediction system that delves into ecosystem and biodiversity indicators can change in the next one to seven months.
What challenge does it address?
Missing the spatio-temporal resolution to capture changes and produce accurate ecosystem predictions, rendering CLMS ecosystem/ biodiversity indicators possess incapable to provide conservation and risk mitigation recommendations.
What’s the breakthrough?
The combination of seasonal forecasts, EO data, Computer Vision Foundation Models, ML/DL, in-situ observations and scientific knowledge to foresee the future of ecosystems.
Why does it matter?
Grasslands and wetlands are increasingly vulnerable to climate pressures. Improved monitoring means earlier detection of changes and timely actions for the protection of these ecosystems belonging to the first resilience pillar.
EO for Ecosystem Health and Degradation
What is it?
An integrated EO approach for monitoring ecosystem health, degradation and changes in habitat conditions.
What challenge does it address?
Existing Earth Observation (EO) indicators have limited ability to capture ecosystem transitions like vegetation and water content, deviations from baseline conditions and biodiversity loss over time. As a result, continuous monitoring of “health” statuses in ecosystems beyond in-situ assessments remains unattainable.
What’s the breakthrough?
It combines long-term EO change indicators, in-situ data and high-resolution products to generate a single health indicator for each habitat type.
Why does it matter?
It enables real-time monitoring of ecosystem conditions and supports faster, timely informed responses to environmental pressures from policymakers.
EO for Small Woody Elements Monitoring
What is it?
An AI-enhanced system for monitoring small woody elements, such as individual trees, within protected ecosystems.
What challenge does it address?
CLMS current mapping is spatially and temporarily scattered, only covering a subset of ecosystems while failing to distinguish between vegetation types.
What’s the breakthrough?
Sentinel-2 imagery is enhanced with AI super-resolution to produce high-resolution maps tracking the gain, loss and growth of woody elements while distinguishing them between broadleaf and needleleaf vegetation.
Why does it matter?
Small wooden elements may be small in size but their contribution to biodiversity and water resilience is significant in impact. Their consistent monitoring can help protect valuable ecosystem features over time.
Using Seasonal Forecasts for Improved Water Management
What is it?
A seasonal water intelligence service providing daily information up to six months ahead for small river sections.
What challenge does it address?
Current droughts and floods are predicted based on weather forecasts or seasonal predictions in the form of monthly averages.
What’s the breakthrough?
Seasonal forecasts are delivered at a local scale with monthly coverage and daily timesteps, looking up to six months ahead, enriched with information on snow, reservoir and upstream storage.
Why does it matter?
Water management becomes easier for the water managers and strengthened with more detailed information on future water availability for seasonal risk assessment and preparedness.
GenAI-Based Hydrological Modelling
What is it?
An AI-driven approach for improving local-scale hydrological and streamflow predictions.
What challenge does it address?
Traditional hydrological models face challenges in representing local factors such as snow, reservoirs, dams and other human-made structures.
What’s the breakthrough?
GenAI/ML learns from observations and model data while incorporating SWE, reservoir and Global Dam Watch information to improve local water-flow estimates.
Why does it matter?
More accurate local streamflow predictions support better water management and preparedness.
Water demand analysis for all the EU
What is it?
A pan-European system for understanding where, when and how much water is needed across different sectors.
What challenge does it address?
The lack of real-time data and an EU information base on water use and demand or comprehensive insights into sectoral water consumption hinder the implementation of the Water Resilience Strategy.
What’s the breakthrough?
It combines Land Use/Land Cover (LULC), agriculture, weather, drought, industry and tourism-related data to generate high-resolution water-demand intelligence from local to pan-European scale.
Why does it matter?
It helps decision-makers balance water demand and availability, supporting smarter planning and investment.
Finding streamflow/in-situ observations with GenAI
What is it?
A GenAI-powered discovery assistant for finding hydrological observations across national data sources.
What challenge does it address?
Streamflow observations are fragmented across countries and institutions and are often difficult to discover, particularly across different languages.
What’s the breakthrough?
GenAI searches national data offers in local languages, helping unlock observations that can enrich open training databases.
Why does it matter?
The expansion of the training network from about 2,000 observation sites to 10,000 stations will improve accuracy and the geographical specificity of the AI stream flow models.
Seasonal Water Resources Stored in Snow
What is it?
A novel Earth Observation operational framework that produces daily estimates of snow depth and SWE (Snow Water Equivalent) in mountainous regions.
What challenge does it address?
Snow-water storage is difficult to measure at the spatial and temporal detail needed to predict future water availability.
What’s the breakthrough?
Earth Observations (EO) are combined with physics-based snowpack models to generate reliable SWE maps at <500 m resolution.
Why does it matter?
Better knowledge of mountain snow storage improves predictions of future river flow and water availability. This breakthrough is also interconnected with innovation 3 in support of the AI seasonal prediction system.
Groundwater (GW) storage prediction
What is it?
A system for predicting and monitoring seasonal groundwater storage across Europe.
What challenge does it address?
Groundwater is difficult to observe directly, while existing EO-based gravity products provide valuable information but at relatively low resolution.
What’s the breakthrough?
Gravity observations from GRACE-FO satellites are combined with hydrological recharge, soil-moisture/SWI information and a ML approach to improve the spatial detail of groundwater storage estimates.
Why does it matter?
It reveals changes in Europe’s hidden water reserves, strengthening long-term water-resource planning.
Compound Energy Drought prediction
What is it?
An AI-powered forecasting system that identifies periods of compound solar and wind shortages.
What challenge does it address?
Extreme climate conditions can reduce solar and wind energy production. Existing forecasting systems rely on historical data and weather models that focus on either solar or wind separately. As a result, compound solar-wind drought forecasts remain limited and difficult to scale.
What’s the breakthrough?
AI builds on existing forecasting approaches and scales them across multiple spatiotemporal levels, from individual PV plants to pan-European systems, while standardising climate-energy risk metrics.
Why does it matter?
It helps anticipate renewable-energy shortages and supports proactive grid resilience and early warning.
ML assessment of desert dust impact on urban and energy sectors
What is it?
An EO and ML framework for assessing how desert dust affects air quality, health and solar energy.
What challenge does it address?
It is difficult to predict dust variability and distinguish its natural contribution from other sources of particulate pollution.
What’s the breakthrough?
New observations from EarthCARE and Sentinel-4 are combined with climate records and ML to quantify dust contributions to total PM and improve future projections.
Why does it matter?
It supports more accurate air-quality assessment, health protection and solar-energy planning.
Assessment of human exposure to heat stress at local scale
What is it?
A high-resolution system for assessing urban thermal comfort and population exposure to heat stress.
What challenge does it address?
City-wide heat indicators can hide major differences between neighbourhoods and population groups.
What’s the breakthrough?
Copernicus data are fused with urban morphology, mobility and vegetation data, surface-flux models and neural networks to produce building-block-level thermal comfort indicators.
Why does it matter?
It helps identify where people are most exposed to heat and supports targeted urban adaptation measures.
AI for mapping Particulate Matters
What is it?
An Earth Observation (EO) and AI framework for producing daily, high-resolution urban PM₂.₅ and PM₁₀ (Particulate Matters) maps.
What challenge does it address?
Satellite PM mapping is often too coarse, while ground stations cannot provide complete coverage across cities.
What’s the breakthrough?
AI combines satellite observations with meteorological, urban morphology and socio-economic data to produce daily, high-resolution PM₂.₅ and PM₁₀ maps at intra-city scale, revealing pollution patterns that conventional satellite products and sparse monitoring stations cannot capture.
Why does it matter?
Air pollution can vary significantly within the same city, putting some neighbourhoods and populations at greater risk. High-resolution PM mapping helps identify pollution hotspots and exposed communities, enabling more targeted action to protect public health.
Monitoring urban CO2 emissions at local scale
What is it?
An EO and AI approach for monitoring urban CO₂ emissions at local scale.
What challenge does it address?
Urban emissions vary across buildings, transport, land use and other sources, making local emission patterns difficult to capture.
What’s the breakthrough?
Copernicus data are combined with urban data on morphology, land cover, mobility and vegetation to train AI models that capture local-scale CO₂ emission patterns with greater spatial and temporal detail.
Why does it matter?
It provides cities with more detailed evidence to monitor emissions and target climate action.