November 3, 2021: xcube - Python package for Earth Observation data cubes

Pangeo Showcase talk by Gunnar Brandt and Norman Fomferra, Brockmann Consult GmbH, Hamburg, Germany

Bio

Gunnar Brandt is a Marine Scientist and holds a Ph.D. in Mathematical Modelling. After some years in research, he joined Brockmann Consult in 2015, where he helped to establish the company’s ARD and data cube activities.

Abstract

xcube is a software package and toolkit for generation and exploitation of analysis-ready data cubes from Earth Observation and other geographical data. It is based on Python’s popular data science technology stack, particularly on xarray, dask, and zarr, and is freely available as open-source software. Data sources for cube generation including the Sentinel Hub service or ESA’s CCI Open Data Portal, are connected via data store plugins. Users can individually configure cube generation by subsetting and resampling input data in space and time or even by applying user code on the inputs. xcube goes further than cube generation by offering high-level methods for data set exploitation and even service provision. Services for cube generation, data provisioning, and particularly cube visualisation via a lightweight web-viewer make xcube a versatile and useful tool for many applications also in operational contexts. In this introductory talk, an overview of xcube features as well as some typical applications are presented.

1 Like

DOI

It’s been a while since this thread began but I recently came across the xcube (group of) software and I must say I’m impressed. I particularly like the “direct” approach of using the standard go-to stack (zarr, dask, xarray) when it comes to interacting with cloud storage services. Way to go!