On-Demand Earth System Data Cubes

Abstract: Advancements in Earth system science have seen a surge in diverse datasets. Earth System Data Cubes (ESDCs) have been introduced to efficiently handle this influx of high-dimensional data. ESDCs offer a structured, intuitive framework for data analysis, organising information within spatio-temporal grids. The structured nature of ESDCs unlocks significant opportunities for AI applications. By providing well-organised data, ESDCs are ideally suited for a wide range of sophisticated AI-driven tasks. An automated framework for creating AI-focused ESDCs with minimal user input could significantly accelerate the generation of task-specific training data. Here we introduce cubo, an open-source Python tool designed for easy generation of AI-focused ESDCs. Utilising collections in SpatioTemporal Asset Catalogs (STAC) that are stored as Cloud Optimised GeoTIFFs (COGs), cubo efficiently creates ESDCs, requiring only central coordinates, spatial resolution, edge size, and time range.

• The paper introduces Cubo as a tool that automates Earth System Data Cube generation by streamlining bounding box calculations and data retrieval.
• It leverages cloud technologies such as STAC and Cloud Optimised GeoTIFFs to efficiently assemble high-dimensional Earth observation datasets.
• Cubo’s standardized workflow supports diverse applications, from global environmental monitoring to disaster response by integrating varied satellite data.

Simplified Generation of AI-focused Earth System Data Cubes Using the Open-source Tool Cubo

Introduction to Cubo and Its Necessity

The concept of Earth System Data Cubes (ESDCs) facilitates structured and efficient analysis of high-dimensional Earth system data. The introduction of the open-source Python tool, cubo, marks a significant enhancement in generating these data cubes specifically optimized for AI applications. Cubo leverages cloud technologies, particularly the use of SpatioTemporal Asset Catalogs (STAC) and Cloud Optimised GeoTIFFs (COGs), to automate and simplify the creation of ESDCs with minimal user input.

Framework and Operational Details of Cubo

The cubo tool simplifies the process of creating AI-focused ESDCs through a streamlined set of parameters and a structured workflow. The user defines only a few critical parameters, including the central coordinates, cube edge size, spatial resolution, and time range. Cubo handles the construction through systematic steps that include bounding box calculation and the retrieval and assembly of relevant Earth observation data into an ESDC. Notably, cubo efficiently transforms spatial coordinates and manages data extraction, adhering to specified spatio-temporal constraints.

Key Steps in ESDC Construction

• Bounding Box Calculation: Utilizes user-input parameters to adjust and calculate the precise bounding coordinates for data extraction.
• Data Retrieval and Assembly: Connects with STAC to fetch relevant datasets, aligning them into a cohesive ESDC format.
• Attribute Annotation: Essential metadata and attributes are embedded into the created ESDC, enhancing usability and integration in downstream applications.

Practical Applications and Demonstrations

Cubo's applicability is demonstrated through two practical scenarios:

1. Global Versatility: Several ESDCs were generated with varying parameters across multiple global locations, showcasing cubo’s adaptability to different geospatial and temporal needs.
2. Standardized Data Synthesis: An extensive, standardized ESDC was created using varied datasets at a single location, demonstrating how diverse Earth observation data can be seamlessly integrated into a single analytical framework.

These examples underline cubo’s potential in facilitating detailed and context-aware analysis of Earth systems, which is crucial for monitoring environmental changes and aiding in disaster response strategies.

Conclusions and Future Implications

Cubo represents a significant progression in the technology available for generating ESDCs, with its capability to substantially reduce the complexity and user effort involved in data preparation for AI applications. The tool’s potential to support advanced analytical tasks and AI-driven research in Earth system science holds promise for future developments, including more nuanced AI models that can predict and interpret complex environmental phenomena.

This contribution is poised to aid researchers in efficiently harnessing the growing volumes of Earth observation data, aligning with current trends towards more integrated and automated data analysis in the geosciences and remote sensing communities.