The Climate Change Science Program Synthesis and Assessment Report 1.1 has documented multiple sources of uncertainty in surface and upper-air temperature records. These error estimates have specific spatial and temporal structures. On global and large-scales, these errors are known to be small such that the climate change detection results are relatively unaffected. However, some biases are known to exist; for example, the record of measurement-type (i.e., specific instruments used) for reconstructing the sea surface temperature fields is known to be incomplete up to the present-day. In this project we are applying a rigorous statistical framework to the analysis of uncertainty in climate model simulation results and global observational data. The effort involves using methods of covariance estimation and optimal climate change detection to investigate: the biases and structures of observational uncertainty, the patterns of variability from climate model results, and the impact of these biases and uncertainty on climate change detection results. The effort makes use of the archive of atmosphere-ocean general circulation model simulations available at the PCMDI IPCC Data Portal for work on the IPCC Fourth Assessment Report. These AOGCM data for the 20th century and pre-industrial control simulations are being used directly in the optimal detection algorithm. We are also developing a hierarchical Bayesian framework for estimating the detection statistics that incorporates the previously identified error sources.