A COMPARISON AND EVALUATION OF TWO CENTENNIAL-SCALE SEA SURFACE TEMPERATURE DATASETS IN THE CHINA SEAS AND THEIR ADJACENT SEA AREAS

Wang Qing-yuan,Li Yan,Li Qing-quan,Wang Ya-nan
DOI: https://doi.org/10.16555/j.1006-8775.2018.04.004
2018-01-01
Journal of Tropical Meteorology
Abstract:Two reconstructed sea surface temperature (SST) datasets (HadISST1 and COBE SST2) with centennial-scale are compared on the SST climate change over the China Seas and their adjacent sea areas. Two independent datasets show consistency in statistically significant trends, with a warming trend of 0.07-0.08 degrees C per decade from 1890 to 2013. However, in shorter epochs (such as 1961-2013 and 1981-2013), HadISST1 exhibits stronger warming rates than those of COBE SST2. Both datasets experienced a sudden decrease in the global hiatus period (1998-2013), but the cooling rate of HadISST1 is lower than that of COBE SST2. These differences are possibly caused by the different observations sources which are incorporated to fill with data-sparse regions since 1982. Different data sources may lead to higher values in HadISST1 from 1981 to 2013 than that in COBE SST2. Meanwhile, the different data sources and bias adjustment before the World War II may also cause the large divergence between COBE SST2 and HadISST1, leading to lower SST from 1891 to 1930. These findings illustrate that the long-term linear trends are broadly similar in the centennial-scale in the China Seas using different datasets. However, there are large uncertainties in the estimate of warming or cooling tendency in the shorter epochs, because there are different data sources, different bias adjustment and interpolation method in different datasets.
What problem does this paper attempt to address?