September 19, 2016
Imagine flying over the surface of a distant planet. Craters and valleys, hills and desert-like plains open up in front of you. Or better still – don’t imagine, just do it! Thanks to a new application called PlanetServer, conceived by Angelo Pio Rossi and colleagues at Jacobs University, this is now easily possible on the Internet. But for the scientists from Bremen, it’s not only about the beautiful views. Multidimensional pictures allow approaching the planets up close and retrieving manifold information at selected locations.
“Despite large technological advances in the last decades, planetary sciences still face similar problems today as in the past: to efficiently process and analyse large-scale datasets”, Rossi explains. PlanetServer provides some remedy. “PlanetServer takes a different approach. It is based on a so-called array-database named rasdaman, a leading one.” Complex datasets with three, four or five dimensions – the arrays – are packed into multidimensional datacubes. The data originate from satellites and sensors constantly orbiting and scanning the planets. In the case of PlanetServer they mainly comprise information about the topography on the Moon and Mars, as well as so-called hyperspectral images: photos depicting many more spectral bands than visible to the human eye. They reveal exciting details about the composition and structure of the planetary surface. The information contained in PlanetServer can be retrieved fast and efficiently through variable queries. Combining and cross-linking different queries enables new insights for the scientists down on Earth.
Is there any ice there? Where can I find a certain mineral based on the way it reflects sunlight? Has water modified rock composition or topography on Mars?
“The queries are done directly in the web browser, it really cannot get much easier than this”, Rossi explains a main advantage of the system. „PlanetServer looks like a website, but it is an application. Each click is like a query for the database.“ The server returns only what is needed to answer each query. „That makes our system much faster and easier to use than many existing applications”, Rossi continues.
Potential users of PlanetServer are geoscientists and the planetary science community that apply the hyperspectral and topographic data for their research. With only a few clicks, they can pose their scientific questions translated into queries on a local, regional or global scale. Moreover, PlanetServer not only reproduces stored data but also links them and reveals complex connections. „The detailed maps are also fascinating for enthusiasts in the public. It’s a bit like Google Earth in space, but it’s all open source, based on NASA’s WorldWind virtual globe“, says Rossi.
The scientists from Jacobs University, including graduated students Ramiro Marco Figuera and Jelmer Oosthoek, started the PlanetServer-project four years ago. It is now available for Mars and the Moon, as for these planets lots of satellite data and images have been collected over the last decades. But the tens of Terabyte of data PlanetServer currently comprises are continuously updated with new research results. And there’s still a lot to be done. In the next months PlanetServer will be expanded to further Solar System bodies. Moreover, Rossi and his colleagues want to collect more archived data, for example to uncover areas not yet imaged. Also, existing data usage shall be improved with more analytical power embedded in the application.
PlanetServer is the Planetary Science Data Service of the EC-funded EarthServer-2 project, supported by the European Unions’ Horizon H2020 program (grant #654367). It tackles the application of so-called Big Data, i.e. very large or complex datasets, in earth sciences. Access to those research results e.g. from satellites or climate models can be challenging for experts, decision makers and the public at large. EarthServer-2 uses latest database and visualisation technologies in order to access and extract meaningful information through standardised queries, following well-recognised geospatial standards as established by the Open Geospatial Consortium.
Additional information at:
Questions will be answered by:
Dr. Angelo Pio Rossi | Professor of Earth and Planetary Science
an.rossi [at] jacobs-university.de | Tel: +49 421 200-3153
