The objective of the GRID-BGC project is to create an end-to-end technological solution for high-end Earth system modeling that will reduce the costs and risks associated with research on the global carbon cycle and its coupling to climate.
Our proposed system consists of a single user-oriented software framework that integrates the following five technology components:
1. A data ingest and interpolation engine that acquires ground-based observations of surface weather as its lowest-level input data and produces high-resolution gridded outputs of surface weather fields, using published and well-tested methods for interpolation between surface observations and extrapolation with elevation in complex terrain (Thornton et al., 1997, Thornton and Running, 1999, Thornton et al., 2000). See an on-line prototype of this technology at http://www.daymet.org
2. A state-of-the-art model of terrestrial carbon, water, and nitrogen cycles that acquires gridded surface weather fields from the interpolation engine, performs a configurable sequence of simulations, and produces a high-volume multi-dimensional gridded output dataset. The high-resolution model (Biome-BGC, v. 4.1.2) deals explicitly with the effects of multiple forcing factors and disturbances on the terrestrial carbon cycle, and has been published and extensively tested (Thornton et al. 2002, Churkina et al. 2002, Law et al. 2001, Law et al. submitted).
3. A post-processing engine that acquires and summarizes the high-resolution biogeochemical model output, evaluates the model results against operational in-situ and remote-sensing observations, and performs spatial scaling analyses against global coupled climate system model outputs. This component also optionally provides summary output for accelerating global model spin-up simulations, and provides information back to the remote sensing systems for use in evaluating high-level remote-sensing driven products (e.g. MODIS net primary productivity).
4. A visualization engine that acquires analyzed or summarized output from the post-processing engine and produces static and dynamic visualizations to assist the user in assessing experimental results, developing new experiments, and effectively conveying high-volume high-resolution model output and model evaluation information to a broad scientific audience. This component will produce output for both desktop display and for high-end visualization environments.
5. A mass storage system with high-speed connection to the computational engines. This is a critical component of the system that solves the problems of data flow and storage in high-resolution, high-volume, rapid-throughput simulation modeling. Our system would make this very expensive technology available to a broad research community by linking remote mass storage facilities, computational engines, and visualization facilities, eliminating the need for dedicated hardware or hardware redundancy. See an on-line description of the National Center for Atmospheric Research (NCAR) Mass Storage System (MSS) at www.scd.ucar.edu/main/mss.html
GRID-BGC Home Page