Canonical Correlation Analysis (CCA) Forecasts of Canadian Temperature and Precipitation -- Apr-May-Jun 1996

In the last two issues of this Bulletin, forecasts of Canadian temperature and precipitation using the multivariate statistical technique of canonical correlation analysis (CCA) were presented. For Canada, we have developed the predictive relationships between evolving large scale patterns of quasi-global sea surface temperature, Northern Hemisphere 500 mb circulation, and the subsequent Canadian surface temperature and precipitation. In this issue we present the forecasts for Apr-Jun 1996 using the predictor fields through February 1995. These forecasts are made with a lead time of 4 months, where lead time is defined as the time between the end of the latest predictor season and the end of the predictand season. Further detail about the Canadian CCA-based seasonal climate prediction is found in Shabbar (1996a, 1996b) and Shabbar and Barnston (1996).

Figure 1(a) shows the CCA-based temperature forecast for the 3 month period of April-June 1996 expressed as standardized anomaly. Table 1 shows the value of the standard deviation in C at selected stations. The mean skill over all 51 stations is given in the caption beneath each forecast map. The field significance is also shown, reflecting the probability of randomly obtaining overall map skill equal to or higher than that which actually occurred. Field significance is evaluated using a Monte Carlo procedure in which the forecast versus observation correspondences are shuffled randomly 1000 times. The field of cross-validated historical skill (correlation) for the forecast time period is shown in Figure 1(b). The forecast has a modest expected skill: a mean national score of 0.14 and a field significance of 0.082. The skill of the temperature forecast drops off considerably in spring in Canada (see the September 1995 issue of this Bulletin, page 28). Local skills are highest over the northern Canadian Prairies, and modest skill is found on the west coast of Canada. A large area of the country from the Rockies to Hudson Bay is expected to experience a negative temperature anomaly; positive temperature anomalies are forecast for both coasts.

Figure 2(a) shows the CCA-based precipitation forecast for the 3 month period of April-June 1996 expressed as a standardized anomaly. Table 1 shows the value of the standard deviation (mm) at selected stations. Cross-validated historical skill (correlation) for this time period is shown in Figure 2(b). The forecast has moderate expected skill: a mean national score of 0.14 and a field significance of 0.050. Local skills are highest over sections of the Prairies, over southwestern British Columbia and the east coast. Large areas of Canada extending from Saskatchewan to Quebec are expected to have a deficit in April-June precipitation. The largest deficit is forecast in central Saskatchewan. An excess in spring precipitation is expected over central regions of British Columbia and the east coast of Canada.

Following the normal evolution of the current cold ENSO episode, some models are projecting a return to normal conditions in the central and eastern equatorial Pacific by the middle of 1996. The April-June '96 forecast recognizes the near future demise of the current cold ENSO episode which started in the middle of last year, and also reflects the decreasing influence of ENSO in the warm half of the year as compared with winter and the first half of spring. Nonetheless, the April-June forecast reflects a strong component of persistence from the winter circulation pattern.

Table 1. Standard deviation of temperature (Temp) and precipitation (Prcp) for the 3 month period April through June at selected Canadian stations.

         	Temp	Prcp

Station      	(oC)     (mm)
Whitehorse	 1.6	 13.2
Fort Smith	 2.5	 19.5
Innujjuak	 1.9	 18.2
Eureka		 2.6	  3.5
Vancouver	 1.3	 26.6
Edmonton	 1.7	 26.3
Regina		 2.1	 30.9
Winnipeg	 2.2	 37.4
Churchill	 2.1	 24.6
Moosonee	 1.9	 27.6
Toronto		 1.6	 30.0
Quebec City 	 1.3	 35.3
Halifax		 1.2	 42.7
St. John's	 1.6	 46.6

Shabbar, A., 1996a: Seasonal prediction of Canadian surface temperature and precipitation by canonical correlation analysis. Proceedings of the 20th Annual Climate Diagnostics Workshop, Seattle, Washington, October 23-27, 1995, in press.

Shabbar, A., 1996b: Seasonal forecast of Canadian surface temperature by canonical correlation analysis. 13th Conference on Porbability and Statistics in the Atmospheric Sciences. American Meteorological Society, San Francisco, California, February 21-23, 339-342.

Shabbar, A. and A.G. Barnston, 1996: Prediction of Canadian seasonal temperature and precipitation using canonical correlation analysis. Mon. Wea. Rev., 124, accepted.

Figures

Figure 1. Panel (a, left): CCA-based temperature forecast for the 3 month mean period of Apr-Jun 1996. Forecasts are represented as standardized anomalies. Panel (b, right): Geographical distribution of cross-validated historical skill for the forecast shown in (a), calculated as a temporal correlation coefficient between forecasts and observations. Areas having forecast skill of 0.30 or higher are considered to have utility. The mean score over 51 stations is 0.14. Field significance is 0.08.

Figure 2. Panel (a, left): As in Figure 1a (CCA anomaly forecast), except for Apr-Jun 1996 precipitation. Panel (b, right): As in Figure 1b (skill for the forecast shown in [a]), except for precipitation. The mean score over 69 stations is 0.14. Field significance is 0.05 (see text).