Flux towers provide temporal quantification of local carbon dynamics at specific sites. The number and distribution of flux towers, however, are generally inadequate to quantify carbon fluxes across a landscape or ecoregion. Thus, scaling up of flux tower measurements through use of algorithms developed from remote sensing and GIS data is needed for spatial extrapolation of carbon fluxes and to identify regional sinks and sources of carbon. Spatial and temporal quantification of carbon dynamics are useful in understanding the biophysical factors that cause regions to be sinks or sources of carbon. We analyzed data sets from the Northern Great Plains and the Kazakh Steppe and found similarities in latitude, precipitation, and carbon fluxes between the two regions. These similarities allowed us to pool carbon flux data, remotely sensed data, and GIS data from these two regions to map gross primary productivity (Pg), total ecosystem respiration (Re), and net ecosystem exchange (NEE) for Kazakh Steppe for 2001 using regression tree techniques. We estimated 10-day Pg and Re with mean absolute errors of 3.2 and 2.7 g CO₂/m²/day, respectively. The NEE for grasslands in the Kazakh Steppe during the growing season (April through October 2001) was 0.79 t C/ha. Localized carbon sinks and sources were positively correlated with growing season precipitation and Pg. The regression tree technique provided an effective method for the regional mapping of carbon dynamics as seasonally quantified by flux towers in the Northern Great Plains of North America and the Kazakh Steppe of Central Asia.