Towards Collaborative Exploration and Analysis of Big Data from Mars: A Noachis Terra Case Study - Towards an Interdisciplinary Approach in Earth System Science - page 228

the visual imagery in a GIS system (Fig. 4 ). This allowed for an integrated in depth
analysis of the possible relation between the hydrated mineral signatures, volcanism
and the (structural) geology.
The results of the Case Study will be presented in a separate publication.

4 Discussion

Mars satellite data are typically processed by individual scientists or by the IT
facilities of their research institutes and universities. Data are shared among
cooperating scientists but are generally not made available online. It is therefore
likely that the same data has been processed by different researchers, independent of
each other.
In the case of CTX data, mosaics used within publications are usually not made
available online. Google, in cooperation with NASA Ames, created a global CTX
mosaic within the Google Earth Mars globe (Andersen 2012 ). Unfortunately this
mosaic is currently not accessible outside Google Earth. In our approach, the
' '

HPC cluster allowed for processing of the CRISM-MRDR-tile-sized
CTX mosaics within a couple of hours. The limiting factor was the maximum
download speed allowed by the PDS. Once all the CTX data of Noachis Terra had
been downloaded the data could be relatively quickly processed, therefore enabling
the explorative phase described in the Case Study. Access to the CTX mosaic tiles
was provided through WMS, enabling the easy sharing of data within our group,
and to potential third parties interested in accessing the data.
In the case of CRISM, processed data are commonly kept of

blackpearl

ine. IDL/ENVI is
used to spectrally analyze the CRISM data and a planetary scientist interested in
connecting the geology (using e.g. CTX) with the mineralogy of a certain region on
Mars will have to go back and forth between a GIS and IDL/ENVI. This limits
CRISM data access to advanced users. The PlanetServer client, on the other hand,
allows users to access and analyze CRISM data online, together with available
overlapping CTX data. The current client has limited WebGIS capability as online
geological mapping cannot be performed. A new client (neo.planetserver.eu,
Chiwome et al. 2014 ), with a more advanced webGIS, is in an early stage of
development and not used within the Case Study.
Therefore ArcGIS in combination with the ArcGIS PlanetServer add-in was
used, allowing for a direct comparison between the geology and mineralogy of a
certain region on Mars in a complete GIS environment. Although IDL/ENVI
remains the most advanced software for spectral analysis, the python add-in can be
easily extended using the many open source libraries of the python programming
language (see for example Laura et al. 2013 ). On the server-side python could also
be used for more computer intensive tasks such as automated mineral detection
algorithms (e.g. Carter et al. 2013a ).
The CRISM data has been processed on PlanetServer itself during early phases,
but later processing is going to be performed on HPC hardware. Although we used

fl