赤道东太平洋地区海温的随机模拟

本文对描述海气相互作用的随机模型作进一步的解释,考虑周期性外力对海气系统的影响,并用这一改进随机进随机动力模型对赤道东太平洋地区的海温作随机模拟,模拟序列的解释方差达到70%。试验表明,考虑周期性外力对海气系统的作用是明显的,它能进一步提高模拟效果,使随机模型更接近实际。 

STOCHASTIC SIMULATION FOR SEA TEMPERATURE IN EASTERN PACIFIC REGION OF EQUATOR

Monthly anomaly series of the sea surface temperature (SST) for the eastern equatorial Pacific are simulated by a stochastic-dynamic model. Composed of second order differential equations and two stochastic forcings, it represents air-sea interactions.The response spectrum of SST is dependent on feedback coefficients determined with a nonlinear least square method. Observed SST series go from Jan-1951 to Dec. 1980 and normalized power spectra are calculated for the observed and simulated series.Using centered differences, a difference equation is derived to define monthly simulated SST anomalies. The correlation coefficient between the simulated and observed series is 0.826 and the explained variance of simulated series is 0.682.Main variation of SST is due to an outer periodic forcing with a period of 12 months from the sun. Periodic forcing is then imposed. There are three experi-ments with the simulation.The effect of the one in whi’ch both air and sea systems have forcing is the bestAnnual periodic forcicing enhances more than 1% of the explained variance.The annual SST series are reconstructed from the simulated monthly series.The correlation coefficient between the simulated and observed annual series is 0.98. It fol-lows that the present model can simulate very well various SST characteristics in the said region.