Horizon Europe · Grant No. 101189962 · M18 Milestone

TERRA Live:
Intelligent Climate Services
from Space

Decisions, Not Pixels.

Fusing Copernicus Earth Observation with AI and Digital Twins to deliver processed climate intelligence, not raw imagery.

ISU WP6 Leadership • Dr. Anwar Sounny-Slitine • International Space University

Audience Poll

A Silent Revolution

Who used GPS today?

Almost every hand in the room goes up. You used it for maps, weather, ridesharing, or routing a package. It is completely integrated into daily life.

Who used Earth Observation today?

Almost no hands go up. Yet, satellite data determined the flood risk of your building, assessed the health of the crops for your lunch, and routed the ship carrying your smartphone.

The Earth Observation Disconnect

We rely on the insights, but we remain disconnected from the data itself. Satellite information is currently locked behind technical barriers, high processing costs, and complex files.

The goal of the TERRA project is to close this gap, translating raw data into immediately actionable decisions for local administrators, coastal managers, and security agencies.

Soil Moisture Chatbot Mockup 🔍 Click to Zoom

The Perspective

The Overview Effect

Astronauts experience a cognitive shift when viewing Earth from orbit: the Overview Effect. They see our planet as a single, fragile, and deeply interconnected ecosystem.

Satellite technology grants us this perspective every day. Earth Observation is a continuous medical check-up for the planet, repeated every few days from 700 kilometers above.

  • Optical sensors capture visible and infrared light, showing changes in vegetation, soil moisture, and water quality
  • Radar (SAR) bounces microwaves off the surface, allowing us to see through clouds, day and night
  • Free historical archives provide 6 to 8 years of baseline records, letting us trace environmental shifts
Earth from Low Earth Orbit (LEO) 🔍 Click to Zoom

Europe's Eyes in the Sky

Copernicus & the Sentinels

Copernicus Logo
ESA Logo

Copernicus is the European Union's Earth Observation programme, the largest in the world. It operates a fleet of Sentinel satellites, each designed for a different mission. The data has been available for 6-8 years, giving us powerful historical baselines.

Sentinel-1 Satellite Sentinel-1 Radar Satellite
S-1

Sentinel-1

C-Band Radar (SAR)
Sees through clouds, day and night. Key for ship detection, oil spill tracking, and mapping changes in coastlines.

Sentinel-2 Satellite Sentinel-2 Optical Satellite
S-2

Sentinel-2

Multispectral Optical
13 spectral bands. Highly accurate for land classification, vegetation health, and inland/coastal water color.

Sentinel-3 Satellite Sentinel-3 Ocean & Land Color
S-3

Sentinel-3

Ocean & Land Color
Wide-swath sensors. Measures sea surface temperature, algal blooms, and coastal water quality parameters.

All Copernicus data is free, open, and publicly accessible. Petabytes of imagery are generated every day.

The Commercial Constellations

New Space: Planet & Maxar

While the European Copernicus fleet provides a comprehensive, free public foundation, commercial New Space constellations add targeted temporal frequency and high spatial detail.

Planet Logo

Planet: Daily Global Coverage

Utilizing a fleet of over 200 Dove CubeSats, Planet images the entire landmass of the Earth every single day.

  • Temporal: Daily revisit rates to spot rapid changes
  • Spatial: Approx. 3 meters per pixel
Maxar Logo

Maxar: Ultra-High Resolution

Deploying large, high-altitude optical satellites, Maxar captures detailed, task-specific imagery of target regions.

  • Spatial: Extreme visual detail (down to 30 cm/pixel)
  • Application: Damage mapping, infrastructure, defense
Resolution Comparison: 3m vs 30cm 🔍 Click to Zoom

Fusing the free, systematic Sentinel data with commercial targeted imagery yields complete environmental intelligence.

Satellite Fundamentals

The Three Dimensions of Resolution

When designing an Earth Observation service, developers must navigate a fundamental trade-off between three types of resolution:

  • Spatial Resolution: The size of the smallest object detectable on the ground. Ranges from 10 meters (Sentinel-2) to 30 centimeters (Maxar).
  • Temporal Resolution: The frequency of passes over the same location. Ranges from weeks, to 5 days (Sentinel-2), to daily (Planet).
  • Spectral Resolution: The number and width of light bands captured. Allows sensors to look beyond visible light to identify water stress, chlorophyll, or soil types.
🧠 Quick Check: Test Your Knowledge

What are the three fundamental dimensions of remote sensing resolution?

Spatial, Temporal, and Spectral Resolution Triangle

From Pixels to Action

Data Processing Levels

Earth Observation data is structured into processing levels, transforming raw instrument measurements into calibrated, atmospherically-corrected observations, and finally into environmental models.

Level 0 & Level 1: Raw & Top of Atmosphere

Level 0 is raw sensor voltage and raw telemetry. Level 1 is geolocated and calibrated (Top of Atmosphere reflectance) but still contains atmospheric distortion and cloud cover.

Level 2: Surface Reflectance

Atmospherically corrected to represent what a sensor would see on the ground (Bottom of Atmosphere reflectance). This is the clean imagery used by researchers for land classification and vegetation analysis.

Level 3: Geophysical Variables

Processed indices and variables derived from Level 2, such as NDVI (vegetation index), NDWI (water index), or Chlorophyll-A concentrations in water.

Level 4: Decision Intelligence (The Last Mile)

The ultimate tier. Fusing Level 3 indicators with historical models and AI to output actionable decisions, e.g., "Close this beach due to an active pollution plume."

Data Processing Levels Diagram 🔍 Click to Zoom
🧠 Quick Check: Test Your Knowledge

Which data processing level represents atmospherically corrected Surface Reflectance?

The Value Gap

100+ Terabytes per Day. Zero Decisions.

Every single day, public and commercial satellite fleets generate over 100 TB of raw Earth Observation data (Copernicus: ~30 TB; Planet: ~20 TB; Maxar: up to 60 TB; Landsat: ~2 TB). But a pixel is not a decision. The vast majority of this data remains functionally locked away from the public administrators, coastal managers, and security agencies who urgently need it.

💻

Too Technical

Raw satellite data requires advanced pre-processing and GIS expertise.

📄

Incompatible Formats

Different sensors produce data in different coordinate systems and file types.

⏱️

Satellites Miss Days

Orbits have gaps. We need data interpolation and modeling to fill the intervals.

🏫

No Integration

Traditional public services lack the tools to ingest raw satellite streams.

The Solutions Layer

Decisions, Not Pixels

The traditional Earth Observation model sells imagery: raw pixels that require highly trained researchers to analyze.

The new frontier sells processed intelligence: clear answers that any decision-maker can understand and act on immediately.

TERRA acts as the intelligent last-mile service layer. It automatically processes, calibrates, and fuses raw data streams, transforming Copernicus pixels into six operational outputs:

  • Pollution Alerts: Automated warnings sent before beach arrival
  • Hazard Maps: Real-time visualizations showing plume movement
  • Vessel Indicators: Flags identifying ships that turned off transponders
  • Erosion Forecasts: 5-year predictions of shoreline retreat
TERRA Intelligent Platform 🔍 Click to Zoom
Pixels
25 TB/day from Sentinel-1, 2, 3, drone fleets, and IoT sensors
↓
TERRA Logo
TERRA Intelligence Engine
AI Analytics, Data Fusion, Digital Twins, and RESTful APIs
↓
Decisions
Automated alerts, local forecasts, maps, and policy briefs

Real Stories

Three Pilots Across Europe

Greece: Water Pollution

Led by University of Thessaly (iPRISM Group). Fuses Sentinel-2/3 with USV drone telemetry to predict agricultural runoff plumes 48 hours in advance in coastal Fthiotida. Explore Demonstrator

Poland: Maritime Security

Led by Blue Dot Solutions. Fuses Sentinel-1 SAR with AIS logs to track "dark vessels" disabling transponders in the Port of Gdansk, monitoring ship traffic and water quality.

Scotland: Coastal Erosion

Uses Sentinel-1/2 optical and radar imagery to run U-Net/LSTM models to extract shoreline vegetation lines, predicting retreat to protect coastal road infrastructure.

TERRA Pilot Visualizations 🔍 Click to Zoom

Impact

From Reactive to Proactive

48h

Decisions before the crisis: contamination alerts arrive two days ahead of ground sensors

98%

AI accuracy in reading coastlines: machines classify what took survey teams weeks

€0

Raw data cost: the pixels are free. The intelligence is the product.

The paradigm shift: traditional monitoring reacts to events after they happen. TERRA's AI-driven intelligence predicts environmental threats before they occur, changing satellite images into active answers.

Advanced Studies Showcase

The Intelligence Layer in Action

ISU students applied the TERRA concept of "Decisions, Not Pixels" to build operational prototypes in just two weeks, translating raw satellite records into intuitive web applications.

🚢 LogiRhein Live

Fuses Rhine water level gauges with historical models to predict shipping bottlenecks 2 to 4 weeks in advance, suggesting multimodal cargo rerouting via an AI assistant.

🌍 ABIL

Maps corporate assets to satellite observations, converting raw Sentinel-2 and 3 thermal and spectral records into a real-time ESG compliance score per facility.

🛒 GiveMeSpace

Builds a centralized, secure space industry marketplace that connects satellite data providers directly with non-expert enterprise customers.

🌊 OceanSentinel

Detects illegal maritime activities by running AI-powered anomaly detection on Sentinel-1 radar, sending automated Telegram alerts to coast guard patrols.

Research Presentation 🔍 Click to Zoom

The Democratization of Space

The GPS Analogy

In the 1990s, GPS was a complex technology requiring specialized receivers and mathematical training. Today, we simply use a navigation app that processes the raw signals in the background and delivers a simple decision: "Turn left in 100 meters."

Earth Observation is currently where GPS was in the 1990s. Planners do not need satellite physics; they need a clean decision dashboard.

🧠 Quick Check: Test Your Knowledge

In which decade was GPS at a similar "expert-only" complexity level as Earth Observation is today?

GPS vs Earth Observation Analogy Graphic 🔍 Click to Zoom

Interactive Workshop

Let's Explore Together: Live Stations

Explore the pilot capabilities interactively as we walk through the three product chains together. Select a station tab below to see Copernicus data in action!

Station Greece · Sperchios River Delta

Goal: Detect agricultural runoff and sediment pollution. Spot the pollution plume where the Sperchios River empties into the Maliakos Gulf.

Interactive Demo

Drag the slider on the map to compare the True Color S-2 image with the Water Quality Index (WQI) showing the nitrate pollution plume in orange.

Sentinel-2 + Sentinel-3 · Fuses WQI · UTH Lead Explore Live Demonstrator
Sperchios River Delta - Greece Water Quality 🔍 High-Res Delta Observations
HIGH TURBIDITY PLUME
S-2 True Color WQI Plume Analysis

Clicking interactive buttons runs real-time HTML/JS sandbox animations modeling actual pilot logic. Explore the full, live demonstrators on the next slide or visit terra-horizon.eu to learn more.

Interactive Demonstration

Live Platform: terra-horizon.eu

The TERRA platform combines Earth Observation, AI, and data-driven analytics for real-world environmental monitoring (water quality, coastal management, and climate hazards). Explore the live portal below or visit terra-horizon.eu to run the demonstrators and learn more.

Thank You

Decisions, Not Pixels.

The new frontier of Earth Observation is not about better cameras or more satellites. It is about turning petabytes of free public data into the specific decisions that protect water, coasts, and communities. That is what TERRA builds.

Consortium: 9 partners across 7 countries.
Project Details: Grant No. 101189962 · Duration: 3 Years (1 Jan 2025 to 31 Dec 2027) · Funding: €1,999,970.45.

TERRA Consortium Partners 🔍 Click to Zoom

TERRA Webinar

Students are welcome to scan this QR code or watch the TERRA Webinar to see the platform and demonstrators in action!

Latest Project Milestones (June 2026)

April 2026

6th Project Board Meeting online. Reviewed demonstrator progress and platform integration.

May 2026

AI Pervasive Systems (iPRISM) lecture at UTH on AI-driven water pollution forecasting.

June 2026

TERRA Live Showcase at Factory 2026 Workshop, hosted at ISU Strasbourg.

🌐 terra-horizon.eu 💬 @horizon_terra ✉️ anwar.sounny@isunet.edu ▶ TERRA YouTube
1 / 16