Summary

This project addresses the evolution of certain geospatial information infrastructures (analysis-ready geospatial data platforms) from a broad perspective, including: the technologies; the data and metadata models; the pipelines and APIs required to integrate those platforms with existing information infrastructures, such as data spaces; and their use in data analysis, decision-making and forecasting activities. And it proposes to tailor those infrastructures, models and tools to advance in the digitalization of a specific economic sector, the agricultural domain, which is not only a large economic sector but also a potential major contributor to a better protection of the environment.

The project is divided into two subprojects because of the broad perspective mentioned in the previous paragraph. The experience of the researchers in one subproject (Analysis-ready geospatial data platforms for the agricultural sector), from Universidad Zaragoza, in geospatial information infrastructures and in the agricultural sector is complemented with the experience of the researchers in the other one (Development of compact storage for analysis-ready geospatial data platforms), from Universidade da Coruña, in the efficient storage and retrieval of data.

This project is linked to the Thematic Priority “8. Alimentación, bioeconomía, recursos naturales, agricultura, clima y medioambiente” (food, bioeconomy, natural resources, agriculture, climate and environment). Geospatial data play a critical role in the digitalization of the agricultural sector. While the use of this data is increasing in this sector, steadily technified in the last decades, it is difficult to keep pace with the increasing volume and variety of geospatial data which is being produced these days. For example, Copernicus, the Earth observation component of the EU space program, generates tens of terabytes per day, all freely available. The project intends to improve on this situation by adapting and extending the concept of analysis-ready data (ARD) for its application to the agricultural domain. It also addresses the necessary technical components and explores the application of agricultural analysis-ready geospatial data to decision-making processes based on machine learning algorithms. The goal is to streamline the incorporation of new geospatial data into decision-making processes, enabling thus farmers to make faster, more informed decisions and to produce more, better food in an environmentally sustainable way.

The main hypothesis of this project is that significant advancements in the digital transformation of the agricultural sector will arise from integrated geospatial data platforms that go beyond the capabilities of the current generation of horizontal platforms. This current generation includes Sentinel Hub, Copernicus DIAS, Google Earth Engine, Microsoft Planetary Computer and others. We propose a shift towards a more vertical approach, where domain knowledge about agriculture is embedded in these platforms. Although the horizontal platforms can be used in agriculture, as in any other sector, this requires a significant initial effort by domain experts, which need to select, combine and pre-process relevant data before proceeding to its analysis. We hypothesize that advancements in vertical, domain-specific platforms will make agriculturally relevant ARD more findable, interoperable, accessible and reusable (FAIR) enabling domain experts to focus only on the problems that they really want to solve. Furthermore, these advancements are expected to be better supported with the integration of Discrete Global Grid Systems in these platforms, especially if we also incorporate the several additional benefits which will arise by replacing the classic uncompressed storage methods with CDS (Compact Data Structures)-based storage methods and their associated query algorithms.

Financiación (Funding)

Proyecto PID2024-155657OB-C21, PID2024-155657OB-C22 financiado por MICIU/AEI/10.13039/501100011033 y por FEDER, UE.