Evaluation And Projection Of The Leading Modes Of Sea Surface Temoerature Anomalies Over The North Pacific In Climate Models

The North Pacific has important impacts on climate in the Northern Hemisphere. It is a key area of global climate change on decadal scales. Studying the leading modes of the sea surface temperature anomalies (SSTA) over North Pacific may help to understand the mechanism by which North Pacific impacts climate. Based on three observed SST data sets and coupled models output, the spatial-temporal characteristics of the leading modes of SSTA over the North Pacific are evaluated and projected in this paper.The spatial-temporal characteristics of the two leading modes of the North Pacific SSTA are investigated for the period1871to2010. The results indicate that the first mode represents well-known Pacific Decadal Oscillation (PDO) pattern with anomalous cooling in the central North Pacific coincident with anomalous warming along the west coast of the United States of America, while the second significant mode exhibits a tripole pattern with northeast-southwest-southeast ’+-+’ SST anomalies. The second mode is referred to as the North Pacific Tripole (NPT). The NPT denotes the inter-annual variability of North Pacific SST, and the NPT variations are closely linked to El Nino Modoki. The strongest correlation between the NPT and the EM occurs when the NPT leads EM by8-9months, indicating that the NPT may be a precursor for EM events.The result of multi-model ensemble mean shows that both the multi-model ensemble mean of CMIP3and CMIP5can simulate the spatial patterns well, but their intensities are lower. The multi-model ensemble mean of CMIP5is improved as compared to CMIP3models in historical runs, but not very large. However, both CMIP3and CMIP5multi-model ensemble means have little abilities to simulate the time series. Although both CMIP3and CMIP5multi-model ensemble mean can simulate the first mode of the North Pacific SST, but the location of the maximum SSTA center lies far in the east coast of Japan to that in observed data sets.The two leading modes will be stable in future under RCP4.5scenario, with slightly enhanced first mode while decreased second mode. However, under RCP8.5scenario, the first mode is a single-signed loading pattern while the second mode shows the spatial variability of the PDO pattern with reverse variability between the central North Pacific and the west coast of North America.