Correlation of spatial climate/weather maps and the advantages of using the Mahalanobis metric in predictions

he skill in predicting spatially varying weather/climate maps depends on the definition of the measure of similarity between the maps. Under the justifiable approximation that the anomaly maps are distributed multinormally, it is shown analytically that the choice of weighting metric, used in defining the anomaly correlation between spatial maps, can change the resulting probability distribution of the correlation coefficient. The estimate of the numbers of degrees of freedom based on the variance of the correlation distribution can vary from unity up to the number of grid points depending on the choice of weighting metric. The (pseudo-) inverse of the sample covariance matrix acts as a special choice for the metric in that it gives a correlation distribution which has minimal kurtosis and maximum dimension. Minimal kurtosis suggests that the average predictive skill might be improved due to the rarer occurrence of troublesome outlier patterns far from the mean state. Maximum dimension has a disadvantage for analogue prediction schemes in that it gives the minimum number of analogue states. This metric also has an advantage in that it allows one to powerfully test the null hypothesis of multinormality by examining the second and third moments of the correlation coefficient which were introduced by Mardia as invariant measures of multivariate kurtosis and skewness. For these reasons, it is suggested that this metric could be usefully employed in the prediction of weather/climate and in fingerprinting anthropogenic climate change. The ideas are illustrated using the bivariate example of the observed monthly mean sea-level pressures at Darwin and Tahitifrom 1866–1995. DOI: 10.1034/j.1600-0870.1997.t01-4-00001.x

[1]  Olga Vechtomova,et al.  Stochastic Gravity , 1999, gr-qc/9902064.

[2]  T. Palmer,et al.  Singular Vectors, Metrics, and Adaptive Observations. , 1998 .

[3]  M. Allen,et al.  Human Influence on the Atmospheric Vertical Temperature Structure: Detection and Observations , 1996, Science.

[4]  W. Thacker,et al.  Climatic Fingerprints, Patterns, and Indices , 1996 .

[5]  W. Thacker,et al.  Climatic indices, principal components, and the Gauss-Markov theorem , 1996 .

[6]  R. Pasmanter Metric structures of inviscid flows , 1996, chao-dyn/9602016.

[7]  Tom M. L. Wigley,et al.  Towards the detection and attribution of an anthropogenic effect on climate , 1995 .

[8]  T. Petroliagis,et al.  Error growth and estimates of predictability from the ECMWF forecasting system , 1995 .

[9]  Z. Tóth,et al.  Degrees of freedom in Northern Hemisphere circulation data , 1995 .

[10]  Klaus Fraedrich,et al.  Estimates of Spatial Degrees of Freedom , 1995 .

[11]  Tim Palmer,et al.  The prospects for seasonal forecasting—A review paper , 1994 .

[12]  G. Meehl,et al.  An example of fingerprint detection of greenhouse climate change , 1994 .

[13]  H. V. D. Dool,et al.  Searching for analogues, how long must we wait? , 1994 .

[14]  K. Hasselmann Optimal Fingerprints for the Detection of Time-dependent Climate Change , 1993 .

[15]  Z. Tóth,et al.  Preferred and Unpreferred Circulation Types in the Northern Hemisphere Wintertime Phase space , 1993 .

[16]  Franco Molteni,et al.  Toward a dynamical understanding of planetary-scale flow regimes. , 1993 .

[17]  Philippe C. Besse PCA stability and choice of dimensionality , 1992 .

[18]  J. Wallace,et al.  Does low-frequency atmospheric variability exhibit regime-like behavior? , 1991 .

[19]  Zoltan Toth,et al.  Circulation Patterns in Phase Space: A Multinormal Distribution? , 1991 .

[20]  W. Elliott,et al.  Evidence for Changes in Southern Oscillation Relationships during the Last 100 Years , 1988 .

[21]  Michael E. Schlesinger,et al.  Detecting changes in global climate induced by greenhouse gases , 1987 .

[22]  Kevin E. Trenberth,et al.  On the evolution of the southern oscillation , 1987 .

[23]  P. Jones,et al.  An Extension of the TahitiDarwin Southern Oscillation Index , 1987 .

[24]  T. Barnett,et al.  Origins and Levels of Monthly and Seasonal Forecast Skill for United States Surface Air Temperatures Determined by Canonical Correlation Analysis , 1987 .

[25]  F. Zwiers,et al.  Statistical Considerations for Climate Experiments. Part II: Multivariate Tests , 1987 .

[26]  Thomas L. Bell,et al.  Theory of Optimal Weighting of Data to Detect Climatic Change. , 1986 .

[27]  R. Chervin,et al.  A Comparison of Month-to-Month Persistence of Anomalies in a General Circulation Model and in the Earth's Atmosphere , 1986 .

[28]  J. Horel Persistence of the 500 mb Height Field during Northern Hemisphere Winter , 1985 .

[29]  Richard W. Katz,et al.  Modeling and Forecasting the Southern Oscillation: A Time-Domain Approach , 1985 .

[30]  P. B. Wright,et al.  The southern oscillation: an ocean-atmosphere feedback system? , 1985 .

[31]  Kevin E. Trenberth,et al.  Signal Versus Noise in the Southern Oscillation , 1984 .

[32]  J. Shukla,et al.  Analogs in the Wintertime 500 mb Height Field , 1984 .

[33]  T. Bell Optimal weighting of data to detect climatic change: Application to the carbon dioxide problem , 1982 .

[34]  Wilbur Y. Chen,et al.  Assessment of Southern Oscillation Sea-Level Pressure Indices , 1982 .

[35]  J. Wallace,et al.  Teleconnections in the Geopotential Height Field during the Northern Hemisphere Winter , 1981 .

[36]  K. Mardia Measures of multivariate skewness and kurtosis with applications , 1970 .

[37]  E. Lorenz Atmospheric Predictability as Revealed by Naturally Occurring Analogues , 1969 .

[38]  E. S. Pearson Biometrika tables for statisticians , 1967 .

[39]  Maurice G. Kendall,et al.  The advanced theory of statistics , 1945 .

[40]  Timothy J. Hoar,et al.  The 1990–1995 El Niño‐Southern Oscillation Event: Longest on Record , 1996 .

[41]  Zoltan Toth,et al.  Intercomparison of Circulation Similarity Measures , 1991 .

[42]  W. Menke Geophysical data analysis : discrete inverse theory , 1984 .

[43]  K. Miyakoda,et al.  Cumulative Results of Extended Forecast Experiments , 1972 .