RELATIONSHIP OF AIR TEMPERATURES IN NEW ZEALAND TO REGIONAL ANOMALIES IN SEA-SURFACE TEMPERATURE AND ATMOSPHERIC CIRCULATION

Abstract

A study has been made of the relationship of New Zealand national-mean air temperature anomalies to regional circulation anomalies (represented by circulation indices), regional average sea-surface temperature anomalies, and the SOI. Partly this was to determine the role of these factors during 1992, which was the coldest year for nearly 50 years. The patterns of large-scale anomalies of regional circulation and SST during the year are described. Correlation analysis was performed on 1950–1993 monthly data filtered into two series to separately represent the monthly and interannual time-scales. National air temperatures and regional mean SST were found to vary in closely similar fashion, with maximum correlation coefficients of 0·60 and 0·87 for the high- and low-frequency series respectively. For the high-frequency series, SST lagged air temperature by half a month and had amplitudes less than half those of air temperature. From an examination of the lags, amplitudes, and maps of pressure and SST anomalies, we conclude, firstly, that the dominant influence on both the air and sea-surface temperatures is the meridional circulation (southerly anomalies resulting in low temperatures), and secondly, that SST anomalies are not a primary cause of national-mean air temperature anomalies at the monthly scale. An oscillation of about 6-month period, which is present in all the high-frequency series, was especially evident during 1991–1993 (with an amplitude of about 1·2°C in air temperature). In the case of the low-frequency series, regional SST lagged national air temperature by 1–2 months and had amplitudes nearly equal to those of air temperature. Both were well correlated with the meridional circulation and the SOI. The 1992 air temperatures were depressed mainly because of the anomalous southerly airflow, which was associated with the prevailing ENSO episode, together with some impact of the Mt Pinatubo eruption. Our results provide good post hoc explanations, but they do not offer much potential for prediction at the monthly-to-seasonal scale, apart from whatever possibilities might result from a better understanding of the quasi-6-month oscillation.