海表面温时间序列的相关性及复杂性研究

摘要

海洋表面温度是海洋热力、动力过程以及海气相互作用的综合结果, 是海洋环境一个重要的参数.为了获得海洋表面温度时空变化的复杂性行为并揭示其内在动力性机制, 根据一个改进的样本熵的方法来分析海表面温时间序列的复杂性;同时使用去趋势波动分析方法研究海表面温时间序列的长记忆性.实验结果表明, 高纬度地区的海表面温变化的复杂度低于低纬度地区.因为高纬度地区海表面温度的季节性更加明显, 序列的规则性更强.相对来说季节性因素对低纬度地区海表面温度的影响不大.不管是长时间序列还是短时间序列高纬度地区的海表面温都具有长记忆性, 且序列的不平稳性很强.但是短时间序列的低纬度地区的海表面温的相关性呈现出多种情况, 序列足够长时, 序列也表现出长记忆性.%Sea surface temperature, as an important parameter of marine environment, is the comprehensive result of ocean thermal and dynamic processes as well as air sea interaction.In order to obtain the complexity of the temporal and spatial variation of ocean surface temperature and reveal its intrinsic dynamic mechanism, the complexity of the sea surface temperature time series is analyzed based on an improved sample entropy method, and the long-term memory of that is studied by the method of detrended fluctuation analysis.The experiment shows that the complexity of sea surface temperature change in high latitudes is lower than that in low latitudes.Because the seasonality of sea surface temperature is more obvious in high latitudes, and the sequence is more regular.Generally speaking, seasonal factors have little effect on sea surface temperature in low latitudes.For long time series or short time series, the sea surface temperature in high latitudes is long-term memory and the sequence is not stable.However, the correlation of sea surface temperature at low latitudes in short time series shows many situations.When the sequence is long enough, the sequence also shows long-term memory.