The objectives of the study were: to evaluate the response of grassland ecosystems to atmospheric change at regional and site scales, and to develop multiscaled modeling systems to relate ecological and atmospheric models with different spatial and temporal resolutions. A menu-driven shell was developed to facilitate use of models at different temporal scales and to facilitate exchange information between models at different temporal scales. A detailed ecosystem model predicted that C{sub 3} temperate grasslands wig respond more strongly to elevated CO{sub 2} than temperate C{sub 4} grasslands in the short-term while a large positive N-PP response was predicted for a C{sub 4} Kenyan grassland. Long-term climate change scenarios produced either decreases or increases in Colorado plant productivity (NPP) depending on rainfall, but uniform increases in N-PP were predicted in Kenya. Elevated CO{sub 2} is likely to have little effect on ecosystem carbon storage in Colorado while it will increase carbon storage in Kenya. A synoptic climate classification processor (SCP) was developed to evaluate results of GCM climate sensitivity experiments. Roughly 80% agreement was achieved with manual classifications. Comparison of lx and 2xCO{sub 2} GCM Simulations revealed relatively small differences.