Multiscale evolution of surface air temperature in the arid region of Northwest China and its linkages to ocean oscillations

The global climate has experienced unprecedented warming in the past century. The multiscale evolution of the warming is studied to better understand the spatial and temporal variation patterns of temperature. In this study, based on the yearly surface air temperature from the gridded CRU TS 3.22 dataset and the ensemble empirical mode decomposition method (EEMD), we investigated the multiscale evolution of temperature variability in the arid region of Northwest China (ARNC) from 1901 to 2013. Furthermore, the possible influences on the ARNC temperature change from the Atlantic Multidecadal Oscillation (AMO), Pacific Decadal Oscillation (PDO), and dipole mode index (DMI) were also discussed. The results indicated that in the past century, the overall temperature in the ARNC has showed a significant non-linear upward trend, and its changes have clearly exhibited an interannual scale (quasi-2–3 and quasi-6–7-year) and an interdecadal scale (quasi-14, quasi-24, and quasi-70-year). Compared with the reconstructed interannual variation, the reconstructed interdecadal variability plays a decisive role in the ARNC warming and reveals the climatic pattern transformation from the cold period to the warm period before and after 1987. Additionally, there were also regional differences in the spatial patterns of change trend in the ARNC temperature at a given time. We also found that the AMO and PDO had significant impacts on the ARNC temperature fluctuation at an interdecadal scale, whereas the DMI had a more important role in warming at the annual scale, which suggests that the importance of oceans cannot be ignored when considering climate change. Our findings deepen the understanding of the temperature changes all over the ARNC in the context of global warming.