A paradigm shift towards decentralized cloud-integrated spatial data infrastructures: Lessons learned and solutions provided for public authorities

Abstract

Digital transformation is a key to turn public authorities into organisations that make decisions based on data-driven insights. The use of big geodata can enable public authorities to tackle complex sustainability issues. However, the efficient management of large amounts of geodata through implementing viable data infrastructures represents a major challenge for public authorities. In this article, we propose a decentralized, cloud-integrated spatial data infrastructure (SDI) to meet the needs of public authorities mandated to provide data services based on earth observation (EO) imagery. We describe the SDI setup and integration of the EO cloud platform \cde, drawing on the specific SDI implementation at a federal, agricultural authority in Germany. Two practical applications are illustrated, underpinning the added value of a cloud-integrated SDI. We elaborate on lessons learned from SDI implementation by summarizing four key findings that may facilitate effective SDI establishment and use, namely i) the need for an organizational strategy, ii) identifying stakeholders, including their participatory roles, iii) long-term financial and human resource planning, and iv) the implementation of a data governance framework. The SDI proposed, serves as blueprint for public authorities aiding them on their path to become providers of data services, leveraging the potential of big geodata, including EO imagery.

Towards optimising the derivation of phenological phases of different crop types over Germany using optical high resolution image time series

“…To avoid such issues, we made use of a new integrated spatial data infrastructure at the Julius Kuehn Institute in Braunschweig, Central Germany. 29 This infrastructure combines access to a substantial amount of earth observation data via the CODE-DE platform (Copernicus Data and Exploitation Platform -Deutschland, www.code-de.org) with a large-scale computing facility and storage environment for the deployment and execution of custom algorithms.…”

Soil-erosion events on arable land are nowcast by machine learning

“…Visually classified aerial images and multidimensional spatiotemporal covariates are used to train and test a random forest. A data cube from the Julius Kühn Institute (JKI) 15 provides Germanywide web-coverage services for the dynamic parameterisation of the model. Panels display the variable importance ranking (A) and confusion matrix (B) for the model trained to detect the occurrence of erosion events on arable land in Bavaria.…”