Carbon Climate E-Science

In collaboration with researchers at the Berkeley Water Center, we're working with scientists from the AmeriFlux collaboration. The Ameriflux collaboration was formed in 1996 to measure the net flux of CO₂ to/from major terrestrial ecosystems. The scientific goal is to understand the small scale processes such as vegetation and soil properties regulating CO₂ exchange and the feedback between those processes and the global carbon cycle interactions with other climate processes. Ameriflux is a subset of the world-wide FLUXNET collaboration.

Our current prototype data server populated with data downloaded from the ORNL Ameriflux site is available off http://bwc.berkeley.edu/Amflux/.

Data analysis and data curation are intertwined. The operations useful for scientific analysis are often also important for understanding data quality. For example, what's wrong in the picture below?

The average temperature should be decreasing with latitude (it's colder at the poles than the equator). What's wrong?

The data is often not reported in the winter. This can affect the science - the measurements must be interpreted with knowledge of the conditional sampling.

Thanks to Gretchen Miller and Deb Agarwal's team for the graphs above.

Presentations:

• Development of the Ameriflux Central Valley Data Portals The BWC proposal for our Microsoft project.
• A View of the Ameriflux Data Download. This deck contains the figures above as well as other simple views of the dataset.
• Early Experience Prototyping a Science Data Server for Environmental Data. A talk given at the Microsoft eScience Workshop at The Johns Hopkins University in October 2006.
• Early Experiences with a Scientific Data Server and the Ameriflux Data. A talk given at the Joint Workshop for WATERS Network Test Beds and CUAHSI Hydrologic Information System (HIS) meeting in November 2006.

Status updates and overviews:

• April 2006 This includes some very simple "how to use databases for the complete novice" advice.
• July 2006 This includes our first web download and early data cube experiences (again, from the complete novice view).
• Ameriflux Scientific Data Server Technology Overview. This talk gives an overview of the bwc.berkeley.edu deployment, our current database schema and some thoughts forward.
• Using the database to rapidly assess FLUXNET data. This talk explores the potential for using simple queries to determining data suitability for cross-site synthesis.