Reduction of Model Systematic Error by Statistical Correction for Dynamical Seasonal Predictions

• Published in: Journal of climate Vol. 12; no. 7; pp. 1974 - 1989
• Main Authors: Feddersen, Henrik, Navarra, Antonio, Ward, M. Neil
• Format: Journal Article
• Language: English
• Published: Boston, MA American Meteorological Society 01.07.1999
• Subjects: Atmospherics; Correlation coefficients; Covariance; Earth, ocean, space; Exact sciences and technology; External geophysics; Geophysics. Techniques, methods, instrumentation and models; Meteorology; Modeling; Oceans; Precipitation; Rain; Sea surface temperature; Seasons; Simulation training; Statistical analysis; Statistical mode; Temperature; Time series; Tropical regions; South America; Brazil; Pacific Ocean; North America; America; Statistical analysis; General circulation models; Bias; Empirical orthogonal function; Numerical simulation; Error correction; Atmospheric precipitation; Singular value decomposition
• ISSN: 0894-8755; 1520-0442
• DOI: 10.1175/1520-0442(1999)012<1974:romseb>2.0.co;2

Singular value decomposition analysis (SVDA) is used to analyze an ensemble of three 34-yr general circulation model (GCM) simulations forced with observed sea surface temperature. It is demonstrated how statistical postprocessing based on the leading SVDA modes of simulated and observed fields, primarily precipitation, can be applied to improve the skill of the simulation. For a given limited prediction region, the GCM has the potential to nonlinearly transform the SST information from around the globe and produce a dynamic solution over the region that can be statistically corrected to account for such features as systematic shifts in the location of anomaly dipoles. This paper does not address the separate question of whether the current generation of GCMs contain information above that which could be extracted using linear statistical relationships with SST. For precipitation, examples are drawn from skillful tropical regions, as well as the moderate-to-low skill Pacific–North American and North Atlantic–European regions. Skill averaged across the analysis domain, as measured by the mean anomaly correlation, is notably improved by the statistical postprocessing in almost all situations where there is at least some real skill in the raw model fields. Postprocessing based on leading canonical correlation analysis (CCA) modes has been compared to postprocessing based on leading SVDA modes. The two methods show small differences, but neither one of the methods can be claimed to do better than the other. A third method, which is based on the leading empirical orthogonal functions of the simulations, has been tested on examples of tropical rainfall where it is shown to also be successful, but with skill generally a little below that based on SVDA or CCA modes. The statistical postprocessing appears to have the greatest potential to improve skill for a variable like precipitation, which can have particularly strong anomaly gradients. Application of the postprocessing to largescale atmospheric fields of 500-hPa geopotential height and sea level pressure produced smaller skill improvements relative to the skill of the raw model output.