大型浅水湖泊与大气之间的动量和水热交换系数——以太湖为例

摘要

湖泊水面与大气之间垂直方向的动量、水汽和热量通量与风速、湿度和温度梯度之间存在比例关系,因此在湖泊水气相互作用研究中,这比例系数(交换系数)是关键因子.在以往的研究中,交换系数通常直接采用水面梯度观测法或海洋大气近地层的参数化方案进行计算本文采用涡度相关系统和小气候系统仪器在太湖平台上直接观测的通量和气象要素,对上述交换系数(最小均方差原则)进行优化,结果为:动量交换系数CDION=1.52×10-1、水汽变换系数CEION=0.82×10-3、热量交换系数CHION=1.02×10-3,与其他内陆湖泊涡度相关观测数据的推导结果一致.本文的研究结果表明:与海洋参数化方案相比,在相同的风速条件下,湖面的空气动力学粗糙度比海洋高,这可能是由于受到水深的影响；如果采用海洋参数化方案,会导致湖泊年蒸发量的估算值偏大40％.太湖的动量、水汽和热量交换系数可以视为常数,可以不考虑稳定度和风速的影响.这是因为本文中83％的数据为近中性条件.敏感性分析表明:如果考虑稳定度的影响,LE模拟值的平均误差降低了0.5 W/m2,H的平均误差降低了0.4 W/m2,u*的计算值没有变化；如果考虑风速的影响,u*模拟值的平均误差降低了 0.004 m/s,LE的平均误差升高了1.3 W/m2,H的模拟结果几乎不受影响.这一结果能为湖气相互作用研究提供参考.%In studies of lake-atmosphere interactions, the fluxes of momentum, water vapor and heat (latent ami sensible heat) are parameterized as being proportional to the differences in Mind, humiclily and air temperature between the water surface and a reference height above the surface. The proportionality or transfer coefficients are often assumed to follow the gradient observation above lake surface or the parameterizations established for the marine atmospheric surface layer. Optimization against the eddy covariance and mieromeieorology measurements made over a large shallow freshwater lake (Lake Taihu) shows that the. Transfer coefficients of momentum ( CDION ) , water ( CEION ) and heat ( CHION ) were CDION = 1.52 ×10 -3, CEION = 0. 82 × 10-3 and CHION = 1. 02 × 10 ~ , respectively. These values are in good agreement with the values derived from the eddy covarianec measurement in other inland lakes. Comparison with oeeanographic parameterizations suggests that lake surfaces were aerodynamic-ally rougher than open oceans under similar wind conditions, which may due to the shallow depth of the lake, and that these parameterizations can bias the annual lake evaporation estimate by as much as 40% higher. Our results also suggest that these coefficients ran be regarded as constants independent of stability and wind speed. Sensitivity analysis indicated that the mean error between calculation and observation of latent heat flux ( LE) decreased 0. 5 W/m2 , that of sensible heat flux (H) decreased 0. 4 W/m2 , and no difference between friction velocity (u ) calculation and observation if stability correction was considered since around 83% of data were in neutral condition. If the effect of wind was considered, the mean error between calculation and observation decreased 0.004 m/s for u, , increased 1. 3 W/m2 for IE, and negligible for H. This study can provide reference for the research on lake-atmosphere interaction.