Principal Investigator (PI): Andrew Bingham, NASA's Jet Propulsion Laboratory
The ability to automatically download only data that meets a predefined need and instantaneously visualize it on a local computer is a concept that has yet to be realized. Our solution of earth science datacasting solves this technology gap. Based on the popular concept of podcasting, which gives listeners the capability to download only those mp3 files that match their preference, earth science datacasting will give users control to download only the Earth science data files that are required for a particular application. In essence, earth science datacasting is a simple, yet powerful informed pull and visualization mechanism. This capability directly addresses the "Data and Information Systems Support for Science Focus Areas and Applications" topic of the Advancing Collaborative Connections for Earth System Science (ACCESS) NRA.
Earth science datacasting will be modeled on the server-client architecture used in podcasting and will leverage existing NASA capabilities. On the server side, the latest data granule is placed in an on-line store and an XML feed is created for the granule. The XML feed is based on the RSS 2.0 standard, with additional namespaces for earth science data. The namespaces are based on the Earth Science Markup Language. In addition, results from analysis and mining the data granule can also be included in the namespace (e.g., information pertaining to the signature of a hurricane or cloud-cover fraction).
On the client side, an RSS 2.0 feed reader monitors the server for new feeds (note: feeds related to different types of data can come from multiple servers). The feed reader will be tuned by the user, via a graphical user interface (GUI), to examine the RSS 2.0 content and initiate a data pull after some criteria are satisfied. The criteria might be, for example, download sea surface temperature data for a particular region that have cloud cover less than 50% and during daylight hours. After the granule is downloaded to the client, the user will have the ability to visualize the data in the GUI.
Initially we will deploy an Earth science datacasting system at NASA's Physical Oceanography Distributed Active Archive Center (PO.DAAC) to provide ocean related satellite data. This provides a good demonstration for illustrating how different types of data (raster, vector and point, respectively) can be visualized. Looking into the future, we believe Earth science datacasting has the potential to significantly increase the utilization of earth science by non-traditional communities. We already see the use of RSS-based technology for delivering news and services to mobile phones and personal digital assistants. We envisage an exciting future where earth science information are also made available in this manner.
For this project we have assembled a team of investigators and collaborators with expertise in earth science data systems, science, applications and outreach, as well as advanced computing and visualization of complex data sets.