Bulk transfer coefficients of momentum and sensible heat over semiarid grassland surface and their parameterization scheme

Abstract

The momentum and heat transfer coefficients used in the atmospheric numerical models usually are taken as constants. In fact, the bulk transfer coefficients change not only with atmospheric stabilities, but also with the air motion and thermodynamic properties of the surface boundary layer. In this paper, the bulk transfer coefficients of momentum and sensible heat are determined by using the data observed by the eddy correlation system and those of the average wind velocity and temperature gradients over the Xilin Guole semiarid grassland in May, 2008. The relations between the bulk transfer coefficients and the gradients in Richardson numbers are analyzed, and the relationship between the bulk transfer coefficients and atmospheric stabilities is also studied. Finally, the changes of momentum bulk transfer coefficients with wind speed and the relations between sensible heat transfer coefficients and temperature are determined. Comparison with the eddy correlation method for observing transfer coefficients, there is a great difference between the computational values of typical Businger-Dyer similarity functions. The revised Monin-Obukhov similarity functions could significantly improve the accuracy of the transfer coefficients. Under the near neutral stratification, there was an obvious interaction between the land surface and the flow condition over semiarid regions. The momentum bulk transfer coefficient varies with the mean wind speed at 10 m according to the quadratic curve laws. Under the low wind velocity condition, the influence of the air flow on the roughness was not significant and the momentum transfer coefficient will increase with increasing wind velocity. While at the high wind speed, the momentum transfer coefficient will decrease with increasing wind velocity, for the air flow changes the roughness height of surface boundary layer. Under the near neutral stratification, there was a quadratic curve law between the temperature and the sensible heat transfer coefficient, which provides a useful parameterization scheme for sensible heat transfer coefficient. The new schemes could parameterize the turbulent fluxes with average wind velocity and temperature gradients data, and does not need to compute the roughness.