Atmospheric concentrations of carbon dioxide (CO2) have increased from approximately 280 ppm (parts per million) to 380 ppm since the beginning of the industrial era. We know that this increase contributes to global climate change. Carbon cycle scientists are trying to understand why plants and oceans are taking up as much carbon as they are, so that they can better predict how these carbon sinks will change in the future. One important way in which scientists gain this understanding is by using atmospheric CO2 concentration measurements, together with information about wind and weather patterns from atmospheric models, to trace back where the releases and uptakes of CO2 are occurring. This information is used together with existing computer models of how plants take up carbon to improve the understanding of what controls variability, in space and time, of carbon sinks.
To understand the carbon cycle, scientists must work with large volumes of data that vary in space and time. This work formats several of these carbon cycle datasets in [KML|http://code.google.com/apis/kml/], a descriptive language for geospatial data, so that that data be explored using [Google Earth|http://earth.google.com/]. Google Earth provides a user-friendly interface for viewing spatial and temporal data that be used by a wide range of user groups, including researchers, teachers, and the general public.
This project was led by [Dr. Tyler A. Erickson|http://people.mtri.org/tyler+erickson/] of the [Michigan Tech Research Institute (MTRI)|http://mtri.org], a research institute of [Michigan Technological University|http://mtu.edu]. The visualization was made possible by the ongoing work of many organizations and researchers working in the area of carbon cycle science. In particular, the CO2 concentration measurements were obtained from [NOAA's Tall Tower Network|http://www.esrl.noaa.gov/gmd/ccgg/towers/] of observing sites and the model results were produced through an ongoing NASA-funded project led by [Dr. Anna M. Michalak|http://www-personal.umich.edu/~amichala/] at the University of Michigan, with several members of her [PUORG research group|http://www-personal.umich.edu/~amichala/Group/] contributing data directly for this application. Additional details on models and datasets are described within the KML file itself.
The KML file is an output from an internet-based information system designed for storing, analyzing, and communicating geospatial and temporal data. MTRI researchers build similar systems, based on open-source geospatial software components, to manage data for a wide variety of government clients. The following open source geospatial packages were used to generate the final KML file:
|| [GeoDjango|http://code.djangoproject.com/wiki/GeoDjango] | Python-based geographic web framework ||
|| [GEOS|http://trac.osgeo.org/geos/] | 2-dimensional geometry modeling and manipulation library ||
|| [proj.4|http://trac.osgeo.org/proj/] | cartographic projections library ||
|| [GDAL/OGR|http://www.gdal.org] | raster/vector translation and processing libraries ||
|| [libkml|http://code.google.com/p/libkml] | KML object generation library ||
|| [numpy / scipy|http://numpy.scipy.org/] | Python scientific computing libraries ||
|| [PostgreSQL|http://www.postgresql.org] / [PostGIS|http://postgis.refractions.net] | open source relational database with geometric object support ||
To learn more about this project, contact [Dr. Tyler A. Erickson|http://people.mtri.org/tyler+erickson/] at [mailto:firstname.lastname@example.org].