Multi-scale Temporal Variation Characteristics of Extreme Temperature in Changchun during 1909−2021 and Its Relationship with Large-Scale Climate Indices

Based on the homogenized daily maximum and minimum temperature records during 1909−2021 in Changchun, the warming characteristics of temperature in Changchun over the recent 100 years were evaluated and the contribution rate of urbanization impact was quantified, the multi-scale variation characteristics of 16 extreme temperature indices were revealed, and then the relationship between extreme temperature indices on different scales and Pacific Decadal Oscillation (PDO) and Atlantic Multidecadal Oscillation (AMO) was further discussed. The results showed that warm indices (SU25, TX90p, TN90p, and WSDI) showed upward trends in fluctuation; while cold indices (FD0, TX10p, TN10p, and CSDI) showed downward trends in fluctuation in recent 113 years. Trends of all indices were at 0.01 or 0.05 significance level, except for SU25, WSDI, and TX90p. The extreme temperature indices in Changchun revealed periodic changes at different scales, which are mainly determined by the first two intrinsic mode functions and the residual signal. Most extreme temperature indices have a 3-year or quasi-3-year main time scale revealing the interannual variations and a quasi-6-year time cycle dominating by decadal variations. Few indices are with significant longer time scales, such as quasi-31-year in SU25, reflecting the characteristics of multi-decadal variability. In the original signal and multi-decade variation, most warm indices (SU25, TX90p, TXx, and WSDI) were significantly negatively correlated with PDO in the same period but significantly positively correlated with AMO. It indicated there are obvious in-phase relationships between warm indices and AMO, revealing the significant modulating effect of AMO on interannual and multi-decade variations on extreme warm indices, but out-phase relationships between them and PDO. While, the situation is opposite for cold indices.