Big Earth Data researchers set a new global standard for earth data grids
A new axis-based data model promises more accurate, flexible, and interoperable Earth observation data across science, policy, and industry
image:
Rasdaman datacube engine supporting efficient storage, querying, and analysis of large multi-dimensional Earth observation datasets across spatial, temporal, and parametric axes, providing a practical implementation of the axis-based grid model for interoperable Earth data services.
view moreCredit: Pebau.grandauer from Openverse | Image Source Link: https://openverse.org/image/6f5a00ff-4650-4867-9774-3956d912c118?q=rasdaman&p=2
\
Earth observation data support critical decisions in climate science, disaster risk reduction, environmental monitoring, and infrastructure planning. These data are typically organized as multidimensional grids that connect measurements to positions in space, time, and other dimensions. While such grid-based representations have been used for decades, their conceptual foundations have remained fragmented. Unclear definitions of grid structure, coordinate handling, and value interpretation have limited interoperability, introduced analytical errors, and complicated the integration of datasets across disciplines and platforms.
As Earth data volumes and complexity continue to grow, extending beyond simple maps and time series into large multidimensional datacubes, these shortcomings have become increasingly problematic. Addressing them requires a fundamental rethinking of how Earth data grids are formally defined.
Now, in a study, published in Big Earth Data on 12 December 2025, Professor Peter Baumann, Professor of Computer Science and Electrical Engineering at Constructor University in Bremen, Germany, reports that a new grid modeling framework resolves many of these long-standing issues. The study introduces an axis-centric approach that redefines how Earth data grids are formally described, modernizing international geospatial standards to better reflect how complex Earth data are produced and analyzed today.
The new framework builds on the updated ISO 19123–1 standard and shifts the focus away from predefined grid types toward independently defined axes. Each axis can represent simple indices, regularly or irregularly spaced coordinates, warped geometries, or algorithmic transformations. By allowing axes to be freely combined, the model can describe a wide range of real-world datasets, from uniform satellite imagery to irregularly sampled climate simulations, within a single, consistent structure.
A key conceptual advance is the clear separation between a grid’s domain and its values. The domain defines where data exist by specifying precise positions in space and time, while values describe what is measured at those positions. This distinction resolves persistent confusion over whether grid elements represent points, areas, or volumes. Instead of embedding such assumptions into the grid itself, the framework treats cells and shapes as visualization constructs layered on top of mathematically well-defined positions.
The study also formalizes how data values are evaluated at arbitrary positions within a grid. Rather than relying on implicit assumptions, the framework defines evaluation using regions of validity, weighting functions, and interpolation methods. This allows irregular sampling, mixed coordinate reference systems, and temporal offsets to be handled rigorously and transparently.
“For decades, scientists have relied on grid definitions that were never designed for today’s data complexity,” Prof. Baumann says. “Our work provides a mathematically sound foundation that finally aligns standards with modern Earth data practice.”
Beyond its conceptual contributions, the framework has direct implications for operational Earth data services. It underpins the evolution of the Coverage Implementation Schema (CIS) 1.1 and the forthcoming ISO 19123–2 standard, which modernize data encodings and support efficient, web-native formats such as JSON. The approach also enables scalable datacube services, including implementations in the rasdaman engine, allowing users to query massive multidimensional datasets across space, time, and additional parameters with high precision and performance.
Prof. Baumann contributes extensive expertise in geospatial standards, data modeling, and large-scale Earth data infrastructures. Together, these efforts position the new grid model as a superset of earlier standards, ensuring backward compatibility while enabling future innovation. Existing datasets and services can continue to operate, while new applications gain the flexibility needed to handle emerging data types and growing data volumes.
“This is about trust in data,” Prof. Baumann explains. “When grids are defined unambiguously, scientists can combine datasets confidently, algorithms behave predictably, and decision-makers can rely on the results.”
Looking ahead, more precise and interoperable Earth data grids could improve climate projections, strengthen early-warning systems for extreme weather, and support evidence-based environmental policy worldwide today. By resolving foundational ambiguities and aligning standards with real-world data practices, this study lays the groundwork for a more coherent, future-proof Earth data ecosystem.
Reference
DOI: https://doi.org/10.1080/20964471.2025.2585732
About Constructor University, Bremen, Germany
Constructor University is a private, international research university located in Bremen, Germany. Known for its interdisciplinary approach and strong global outlook, the university brings together students and researchers from over 100 countries. Constructor University focuses on cutting-edge research in science, engineering, and data-driven disciplines, with a strong emphasis on innovation, societal impact, and real-world problem solving. Through close collaboration with industry and international partners, the university contributes to advancing knowledge and developing solutions for global challenges.
Website: https://constructor.university/
About Professor Peter Baumann from Constructor University
Peter Baumann is a Professor of Computer Science and an entrepreneur with internationally recognized expertise in large-scale data management. At Constructor University, Germany, he leads research on flexible and scalable datacube services and their applications across science and engineering. He is the pioneer behind the rasdaman engine, through which he and his team introduced datacubes and Array Databases, establishing the de facto standard for multidimensional data services. His work is documented in more than 200 scientific publications, supported by international patents, and has received numerous high-ranking innovation awards for its impact on data-driven science.
Funding information
Work in part was supported by EU Horizon (StandICT grant 101091933), North Atlantic Treaty Organization G5970, HORIZON EUROPE Framework Programme 09-1220.
New axis-based grid model illustrating how independent spatial, temporal, and parametric axes combine to represent complex Earth observation datasets, enabling consistent alignment, interpolation, and scalable analysis across heterogeneous data sources.
Credit
Peter Baumann from Constructor University, Germany | Image Source Link: https://www.tandfonline.com/doi/full/10.1080/20964471.2025.2585732#d1e160
Journal
Big Earth Data
Method of Research
Computational simulation/modeling
Subject of Research
Not applicable
Article Title
An introduction to the OGC/ISO coverage and datacube standard for modeling multi-dimensional, spatio-temporal Big Data
No comments:
Post a Comment