The Assimila data cube
The Assimila data cube is opening up a wealth of Earth observation and environmental data.
The environment is being observed and analysed like never before-with xx petabytes of data being collected daily. Weather, climate, vegetation, pollution, soils, oceans and the land surface are being measured in unprecedented detail.
Sophisticated modelling techniques are also providing scenarios of future conditions.
However, thevolume and complexity of such data are a barrier to realising their full potentialfor a vast range of applications.
The Assimiladata cube overcomes these barriers by providing easy access to Earth observation data and analysis tools which, in partnership with clients, we can customise for particular applications.
The cube is a tool that works for you –whether you are a University research group or multi-national business.
What does the Assimila data cube provide?
- Access -a set of tools that provide access to multiple data portals to either directly download or access data in a cloud environment.
- Pre-processing –to provide analysis-ready datasets with user-specified resolutions, projections and formats.
- Analysis -smart spatio-temporal querying and analysis tools tailored to your needs.
The technical basis
The core of the Assimila data cube is Cloud Optimized GeoTiffs (COGs) to enable efficient workflows on the cloud. Several Python libraries such as xarray are used to manage the multi-dimensional data arrays, since real world datasets are more than rows and columns. X-arrays have labels to encode locations in space and time. Additionally, we use DASK for parallel computing in Python.
Data is downloaded automatically and catalogued in a PostGIS database. We have a layered object-oriented(OO) architecture which can then deliver these data via an http interface. Users can request timeseries for locations ranging from one point to the whole globe, via a Python API, through Jupyter Notebooks or via a QGIS plugin.
How is the Assimila Data Cube being used?
The Assimila data cube is being used to provide easy access to key Earth observationdatasets that allow the characterisationand assessment over time of peatland conditions, e.g. to assess vegetation recovery after a fire has occurred.We are working with the Universities of Reading and Durham.
See a Jupyter Notebook demonstration.