Methodology for the Qualitative and Quantitative Analysis of the Environmental Impact of Copernicus Data Space Ecosystem (CDSE) End-Users

The Copernicus Data Space Ecosystem (CDSE) has transformed Earth Observation data access cloud-based, multi-access services, replacing traditional download-heavy approaches. By 2024, CDSE provided over 78 PB of data to more than 290,000 users spanning scientific, institutional, and commercial sectors. This rapid expansion, however, raises sustainability questions due to the environmental impact of data transmission, storage, and processing. This study evaluates end-user contributions to the CDSE's environmental impact and explores strategies to reduce emissions across the value chain.

The approach combines qualitative and quantitative analysis across 3 phases:

Phase 1 – Preparation and Validation

The first phase involves defining the scope of the study, validating user typologies (scientists, industry, institutions, start-ups), confirming methodological standards, and developing tools for surveys and interviews.

Phase 2 – Data collection and user insights: Data will be gathered through 3 complementary channels:

• A GDPR-compliant questionnaire targeting diverse typologies
• In-depth discussions to capture decision-making processes and sustainability trade-offs.
• Existing data sources: Bibliographic research, statistical reports, and operational data from CDSE platforms to quantify download volumes, compute operations, and storage patterns.

Phase 3 – Impact calculation and modelling

Impacts will be evaluated for each CDSE end-user type by examining 3 main areas: data transfer (network traffic), data processing (computing tasks), and data storage (including backup and retention). Both cloud hosting and on-premises systems will be analyzed.

The calculation process commences with the collection of key inputs, which include:

• Cloud resource consumption: Data from CDSE operations and cloud providers (compute, storage, data transfer).
• Technical specification of instances used: CPU, GPU, memory, storage type, PUE, etc.
• User behavior data: Collected via surveys and interviews, considering the amount of data, the types of processing performed, how long data is stored, redundancy measures, and involvement of third parties.

Our methodology is adapted from the Cloud Carbon Footprint approach and established standards for assessing the environmental impact of cloud services and data centers, tailored specifically for Copernicus. This evaluation covers 2 emissions:

• Embodied Emissions related to manufacturing and maintaining servers and storage devices

Manufacturing emissions × (usage time ÷ lifespan) × (reserved resources ÷ available resources).

• Operational Emissions caused by resource use during data processing.

Time × total component power × efficiency × electricity carbon intensity.

The calculation uses quantitative data, standard guidelines, established methodologies, and databases such as EcoInvent, EEA, Negaoctet, and Resilio.

Complementing this, a multicriteria LCA approach, following European guidelines and standards, offers a comprehensive view. Key indicators considered are electricity consumption, GHG emissions, primary energy consumption (to capture total energy demand), water consumption (linked to cooling and infrastructure) and abiotic resource depletion (impact of raw material extraction for hardware).

To achieve representative results, sampled typologies are extrapolated to all users using factors such as average footprint, CDSE registration data, infrastructure location, and storage and processing scenarios.

The study will provide:

• Carbon footprint for each end-user type
• Hotspot identification in user types and infrastructure
• Emission reduction recommendations