Scientists Collaborate Worldwide to Gather Critical Data on Climate Change
Managing our terrestrial resources effectively in the face of climate change is a challenge that is reaching critical mass, and needs to be fully explored. In 2003, plants across North America absorbed approximately 30% of our fossil fuel emissions from the atmosphere. Younger forests that were cut decades ago are responsible for about half of that. As these forests age, their ability to absorb CO2 is expected to decline. How much decline and how fast that decline might be is unclear since the forests interact with the changing climate.
Carbon-climate scientists focus on these kinds of interactions between climate and the terrestrial biosphere. Over the past year, scientists at Max Planck Institute for Biogeochemistry in Germany, University of Tuscia in Italy, Oak Ridge National Laboratory, and University of California at Berkeley have teamed with computer scientists at the Berkeley Water Center, University of Virginia, and Microsoft Research to produce a world-wide carbon-climate field dataset. This dataset is now being used by over 200 researchers to produce synthesis results across the field sites as well as integrate with climate models.
Fluxnet is a "network of networks" – an umbrella organization for existing carbon-climate collaborations such as CarboeuropeIP, AmeriFlux, or Fluxnet-Canada. In February 2007, at the La Thuile workshop, scientists began to assemble a world-wide carbon-climate field dataset. The dataset originally compiled at the La Thuile workshop contained approximately 600 site-years of data. Since the workshop, the data set has grown to over 950 site-years from over 250 sites in 35 countries. This is now a living dataset - expanding each year as new data are taken and new sites are deployed. Associated with the time-series data are a number of site properties such as last time of fire or type of vegetation important for the synthesis analyses.
The dataset is currently being used by over 60 paper-writing teams. The American or AmeriFlux subset is also being used by scientists in the North American Carbon Program to integrate with other carbon cycle data sources and models. Example questions include:
The size, complexity, and dynamic nature of the dataset meant that a simple flat file repository was no longer enough. The computer scientists are levering SQL Server databases and data cubes to curate and host the dataset for the scientists. The external face to the science community is http://www.fluxdata.org. The site consists of a Sharepoint collaborative portal paired with a scientific data server which holds the archive SQL database containing the dataset, and a collection of SQL Server Analysis Services data cubes that enable simple data browsing and summary report generation.
The combined servers enable browsing of the data on-line, data download, and version tracking. There are also a number of data summary products such as annual values with variance. The data cubes provide organization and aggregation of data along dimensions such as time to allow easy retrieval of values sorted by site biome, latitude band or other characteristic.
The site also provides collaborative space for scientists on paper-writing teams and the field scientists contributing the data. Scientists can track data versions, report problems with the data, and get updates to reported site properties. The site also helps researchers using the data to communicate with the scientists providing the data.
Authorized members of the Fluxnet collaboration can sign in at the web site and obtain access to a range of additional functions. The "Data Browsing" tab provides access to all available site ancillary data, summary annual values for the variables, and direct connection to the SQL data cube for browsing of the data. The "Data Download" tab enables the scientist to select download of CSV or MatLab format site-year data files. Changes to site ancillary information can also be submitted via the "Ancillary Data Update" tab. Proposal coordinators are able to maintain the list of sites in use for their analysis and communicate with site PIs via e-mail.
Using the Fluxdata web site and Scientific Data Server, researchers can keep up-to-date with the latest updates to the dataset, select sites of interest for analysis, and communicate with others in the collaboration. The resulting site and capabilities we believe allow individual researchers to concentrate on science rather than data management.