Correlations between North Atlantic Oscillation Index in winter and eastern China Flood/Drought Index in summer in the last 530 years

Abstract

The time series of winter North Atlantic Oscillation Index (NAOI) in the period of 1429–1983 developed by Glueck and summer Flood/Drought Index (FDI) of eastern China in the period of 1470–1999 from 100 stations are used in this paper to study the potential impact of North Atlantic Oscillation on the climate in China. The analysis has explored some significant lag correlations between FDI and NAOL The maximum positive correlation coefficients between NAOI and area-mean FDI in eastern and northern China lagging 2–3 years reach at 0.001 significance level, and while there are also negative correlation between NAOI and FDI in central and southern China at significance level of 0.05–0.01. The correlation between FDI and NAOI is time-dependent, i.e. the correlation coefficients between two indices vary from period to period. The highest correlation appeared in the period of 1636–1742, around the Little Ice Age, with the significant level of far above 0.001. The second significant period was from 1951 to 1999, at the level of 0.005–0.002. Both the power spectrum analysis and Morlet wavelet transformation have presented an interesting phenomenon: the area-mean FDIs in eastern and northern China share almost the same oscillation periods with NAOI in the inter-annual, decadal and centurial scales’ oscillations, i.e. 4–5, about 10, 20–30, around 50 and 80–100 years, etc. The Mann-Kendall Rank Statistic test reveals the significant trend and decadal abrupt changes in the series of area-mean FDIs in eastern and northern China in the past 530 years, while the NAOI in the past 400 years, did not show such trend at the significance level, but presented more frequent changes than those of FDI in China. This difference is perhaps due to the fact that the amplitude of the extremes of reconstructed NAOI series is less than that from instrumental records.