The Tropospheric Emission Spectrometer (TES) on board NASA's Aura satellite (to be launched in 2004) will provide measurements of global distributions of ozone, CO, and other key chemical species in the troposphere. In order for TES to meet a design lifetime of 5 years, it has been determined that a global survey strategy with ~50% duty cycle must be identified. In this study, simulated concentrations of ozone and CO from the GEOS-CHEM global three-dimensional (3-D) model of troposphere chemistry are used as a time-varying synthetic atmosphere for demonstrating and assessing the capabilities of TES nadir observations. Autocorrelation analyses of the model species fields for different time lags identify a significant 1-day correlation and support a 1-day-on/1-day-off observation strategy. Three major steps are then taken to demonstrate and evaluate TES products: (1) species profiles along the TES orbit track are sampled from the model 3-D time-varying fields with cloudy scenes (50-60% of total scenes) removed; (2) nadir-retrieved profiles (level 2 products) are obtained from these true synthetic profiles using TES retrieval characteristic functions; (3) interpolated daily global maps (level 3 products) are generated to compare with the original model fields. The latter comparison indicates that the error in the level 3 products relative to the true fields for ozone and CO is <10% in ~70% of cases and <20% in 80--90% of cases. The three major sources of error lie in the asynoptic orbital sampling, the retrieval, and the level 3 global mapping.