摘要
利用中国科学院大气物理研究所大气科学和地球流体力学数值模拟国家重点实验室发展的气候海洋模式(LICOM),考察了两种热通量边界条件(牛顿冷却型边界条件和总体公式型边界条件)对热带太平洋海温平均态和年际变率模拟效果的影响。结果显示,在两种边界条件下,模式均能较好的再现海温的年平均空间分布特征和季节循环特征。对比分析发现,在牛顿冷却条件下,模拟结果与观测更加接近,这是因为该条件会通过调整净海表热通量使模拟海温向观测的气候态海温逼近。就年际变率而言,牛顿冷却条件下模式模拟的净海表热通量负反馈作用偏强,从而使ENSO模拟偏弱,进而使中东太平洋的异常经向温度平流模拟偏弱,造成海温异常的经向尺度偏窄。负反馈的强度与耦合系数的选取有关。而总体公式条件下模式能够合理地模拟出ENSO相关的热通量负反馈过程,从而能正确的模拟出ENSO振幅以及ENSO空间型。因此,当利用海洋模式对气候平均态海温进行模拟时,两种条件均可采用,但以牛顿冷却条件为佳;而当对海温的年际变率进行模拟时,应该采用总体公式型边界条件。
Both Newton cooling scheme(i.e.prescribed total heat fluxes added with a relaxation term) and bulk formula parameterization scheme are widely used in offline Ocean General Circulation model simulations.These two thermal forcing schemes are compared here in two simulations using LASG/IAP climate ocean model(named LICOM) in terms of the performance over tropical Pacific Ocean.The results indicate that for both the annual mean SST distribution and the annual cycle of SST over tropical pacific,the two schemes show reasonable performance,although the Newton cooling scheme shows even less bias due to its relaxation to observed climatological SST.The El Nio amplitude in the Newton cooling scheme simulation is weaker than that in observation,due to the overestimated negative heat flux feedback which is related with the thermal coupling coefficient.And the suppressed El Nio could further lead to a narrower meridional scale of the SST anomalies in the eastern Pacific.With a reasonable negative heat flux feedback in bulk formula simulation,the model exhibits a more reliable El Nio amplitude and spatial distribution.The study demonstrates that if the simulation aims to get a realistic mean state,the Newton cooling is recommended,but if the simulation focuses on interannual variability of the tropical ocean,the bulk parameterization scheme should be used.
出处
《海洋学报》
CAS
CSCD
北大核心
2011年第4期9-18,共10页
基金
国家自然科学基金(40890054)
国家重点基础研究专项经费基金(2010CB951904)