Multiscale characteristics of Jing-Jin-Ji’s seasonal precipitation and their teleconnection with large-scale climate indices

Abstract

Using wavelet theory, Mann-Kendall non-parametric test, principal component analysis, and correlation analysis, the study investigated the change characteristics and periodic variation of seasonal precipitation and their relationships with seven large-scale global climate indices including Nino 3.4, Southern Oscillation Index (SOI), Pacific Decadal Oscillation (PDO), Pacific North American Index (PNA), North Atlantic Oscillation (NAO), Arctic Oscillation (AO), and Eastern Asia/Western Russia (EA/WR), for the period of 1959–2016 in Jing-Jin-Ji metropolitan region. The results showed the following: (1) The change characteristics of Jing-Jin-Ji’s precipitation varied for different seasons. Only JJA precipitation showed statistically significant decreasing trend at 5% significance level. Precipitation in other three seasons had increasing trends. The trend of annual precipitation was mainly dominated by JJA precipitation. The results of fractal characteristics of precipitation indicated that the future trend of JJA precipitation would likely increase, but the future trends of annual precipitation and precipitation in other three seasons would likely decrease. The inapparent periodic variation of seasonal precipitation was revealed using adjacent-average smoothing method. (2) Wavelet theory was used to investigate the periodic variation of seasonal precipitation and seven large-scale climate indices. In general, the oscillations varied for different seasons and periods. The real part of the Morlet wavelet amplitude exhibited similar cycles using wavelet power spectrum. However, it is possible to estimate the future trends of seasonal precipitation according to the nested periodic positive and negative cycles. The periods of seven large-scale climate indices were scattered. (3) The PC1 of seasonal precipitation and climate indices had isolated coherence, which varied for different climate indices and periods. The phases between them also varied for different periods. As for different climate indices, the EA/WR had strong positive impacts on the seasonal precipitation at relatively large period, such as 8–16 years. The effects of climate indices on the seasonal precipitation varied for different PCs and climate indices. However, the results of correlation analysis were consistent with those of wavelet coherence and phase difference. The corrections varied for different time lags between climate indices and seasonal precipitation. The results should be beneficial to reveal the historical and future trends of seasonal precipitation in Jing-Jin-Ji metropolitan region. The linear and nonlinear relationships between seasonal precipitation and climate indices indicated that it might be possible to predict the seasonal precipitation using the climate indices as potential predictors. However, the mechanisms of climate change impacts on the precipitation need to be further elucidated, which are recommended in our future study.