Discussions

EarthServer-2: Agile Analytics on Big Data Cubes

  • Angelo Rossi profile
    Angelo Rossi
    2 November 2015 - updated 4 years ago
  • 1 comment
    Total votes: 3

Earth (and Planetary) Remote Sensing, weather, climate data and assimilation/model results consist of or produce a vast amount of datasets, either co-located or, more commonly, scattered among data centres and not readily (or at all) usable, for scientific use, particularly if in combination.

The core idea of asking "questions" to those data in terms of coherent queries is one of the aims of EarthServer, in its FP7 and H2020 phases: The EarthServer answer to this is to use high-level, standards-based query languages which unify data and metadata search in a simple, yet powerful way.

Phase 1 (FP7) of EarthServer has advanced scalable array database technology into 100+ TB services. Those efforts were witnessed by independent experts who attested that "with no doubt" EarthServer "has been shaping the Big Earth Data landscape". 

In phase 2 (H2020), Petabyte multidimensional datacubes will be built across Europe and Australia to perform ad-hoc querying and merging, without any physical storage colocation need. 

Returned results, compliant with Open Geospatial Consortium (OGC) standards could potentially be used as inputs for scientific exploitation within other running or started projects.

EarthServer has established and is further advancing client and server technology for spatio-temporal datacubes. The underlying scalable engine is the rasdaman Array Database, enabling interaction, data processing, and overall complex analytics. 

For example, data already available include Earth Observation (e.g. http://eodataservice.org/mea/ one of EarthServer-2 data services) queried for surface characteristics/spectra and their temporal change over very long time series can be valuable both for scientific analyses and decision-making processes.

There are several potential opportunities for cooperation across project that we hope to explore and develop further.

More info and updates on http://earthserver.eu/

Angelo Pio Rossi & Peter Baumann

Tags: 
Earth ScienceRemote SensingEarth ObservationenvironmentweatherclimateLandocean

Comments

Log in to post comments
Antonios Barbas's picture
Antonios Barbas,
03/11/2015 12:08

There appears to be a certain level of hesitation to initiate discussions within the Project Group on Earth sciences, energy and environment. Indeed, the projects are very heterogeneous, they stem from different calls and they also have different starting dates.

There are few scientific domains that may be of interest to some of the projects –such as modelling and simulation –but again the exact work orientation and targets are different.

From an e-infrastructure perspective, there may be interest to explore synergies in IT-related issues including interoperability, data analytics, repositories, citation, preservation, processing/computing needs, tools and services.

While at our upcoming concertation first discussions may help to find common ground, I would like to point to the fact that most project DoAs include tasks related to: User/communities and stakeholder engagement, Use cases/demonstrations, Dissemination/showcases, Sustainability exploitation/take-up/business plans, Web platforms, Provision of services, Service and resource commons, Innovation management, Impacts, etc.

These issues are likely to become more important with the evolution of the work, but could be considered from now for potential synergies and cooperation.

Total votes: 1