Abstract

NWP type forecasts made with climate models initialized from reanalyses have proven to be an effective method to diagnose parameterization errors (Phillips et al. 2006, Boyle et al. 2005, Williamson et al. 2005, Williamson and Olson 2007). This approach is most powerful when paired with field cam... Show moreNWP type forecasts made with climate models initialized from reanalyses have proven to be an effective method to diagnose parameterization errors (Phillips et al. 2006, Boyle et al. 2005, Williamson et al. 2005, Williamson and Olson 2007). This approach is most powerful when paired with field campaign observations of parameterized variables. However the approach can also be applied to examine errors and to test hypotheses associated with those errors by using the reanalyses for verification as well initial conditions. There is an advantage in using several reanalyses for the initial conditions. The reanalyses are good enough in many regions, but not necessarily all. Model parameterizations still dominate the analyses in some regions for some phenomena. If the forecasts exhibit the same errors when initialized from and verified against different reanalyses we have more confidence that we are looking at a model error, rather than the difference between the model and an analysis model. Show less