CYBERINFRASTRUCTURE FOR PLATE TECTONIC RECONSTRUCTIONS

Synthesis of diverse geological and geophysical data with plate tectonic reconstructions is becoming increasingly important. Present day locations of data from paleobiology, geochemistry, and stratigraphy, all can be reconstructed to paleo-positions. Visualizing how this information evolves across time and geography can help to understand connections between the observations. Moreover, reconstructions are becoming increasingly important as inputs into solid earth, and long-term climate change models. However, researchers seeking to combine data from multiple sources, or specialists in a particular field, may not have the expertise to effectively use tectonic models. Given the diversity of geo data formats, researchers may not have the resources to efficiently recast data for combining and visualizing.

We have created the GPlates software tools to help solve these problems. GPlates is primarily an interactive desktop client used to create global plate models, and visualize and integrate geophysical data. Unlike a traditional GIS, GPlates is able to take present day geographic information and reconstruct the data to paleo-positions. We have expanded the integration abilities of GPlates by enabling access to remote data via web services (e.g., Macrostrat). Users can combine data on local disks with remotely accessible data. As part of the NSF EarthCube, we are building connections with large centers (IRIS, UNAVCO), as well as research-oriented projects (PaleoDB).

We have also developed server-side solutions to provide simple to use tools to reconstruct data. We have expanded the GPlates software package to include command-line tools and a Python API. From these interfaces we have built a set of web services using the GPlates reconstruction engine. We use common open source technologies and formats to ensure opportunities to the entire earth science community. Our native data format is based on GML and GeoSciML, and we utilize other standards like GeoJSON and ESRI's shapefile. We have built the GPlates web services using a combination of open source software tools (Python, Django) and geography projects (GMT, GDAL).