南海西南季风期NCEP2湍流热通量的质量分析

Evaluation of daily NCEP2 turbulent heat fluxes during southwest monsoon in South China Sea

以5次南海现场观测试验数据(Xisha2002,Xisha2000,Xisha1998,Kexue 1和Shiyan 3)为参照,对NCEP2再分析资料中湍流热通量在南海西南季风期的精度进行了评估.结果表明NCEP2估算的潜热通量的平均值在试验Xisha2000,Xisha1998,Kexue 1和Shiyan 3期间分别高估了6(11%),2(2%),7(7%)和13W/m2(16%),而在Xisha2002试验中低估了10 W/m2(11%).在5个试验中低估的感热通量分别为7(130%),3(64%),7(170%),5(53%)和5 W/m2(72%).NCEP2与5个现场观测试验的时间序列的相关系数均没有达到95%的置信度.模式中湍流热通量损失的误差来源于基本变量和算法,基本变量中以海表温度和海面风速的误差产生的影响最大.应用COARE2.6a算法和NCEP2的基本变量重新计算的湍流热通量更加符合物理意义.

The commonly used turbulent heat fluxes in the National Centers for Environment PredictionNational Center for Atmospheric Research reanalysis 2 （NCEP2） are evaluated using three air-sea flux research tower observations and two cruises in the South China Sea Monsoon experiment （SCSMEX）as the references. The mean latent heat fluxes from NCEP2 are overestimated 6 （11%）, 2 （2%）, 7 （7%） and 13 W/m^2 （16%） during experiment Xisha2000, Xisha1998, Kexue 1 and Shiyan 3 respectively and underestimated 10 W/m^2 （11%） in Xisha2002. The overestimations in sensible heat flux change from 7 （130%）, 3（64%）, 7（170%）, 5（53%） to 5 W/m^2（72%） in Xisha2002, Xisha2000, Xisha1998, Kexue 1 and Shiyan 3 experiments. The correlations between the time series of NCEP2 and the observations are relatively weak. The errors from the oceanic turbulent heat losses in NCEP2 are mainly due to bulk variables and algorithm. In all variables, the errors coming from sea surface temperature and wind speed are especially important. The results suggest that the recalculated NCEP2 surface turbulent heat fluxes using COARE algorithm are more physically reasonable, especially for latent heat flux.