Turbulence decay in the stable Arctic boundary layer

This paper focuses on the study of turbulent transfer in the atmospheric boundary layer under very stable conditions. Results are based on the data collected at a 20-m tower over the Arctic ice during the SHEBA. As stability increases, turbulence decays and vertical fluxes vanish. However behavior of turbulent fluxes and other characteristics including determination of the critical Richardson number are poorly understood in the very stable conditions. We consider behavior of the momentum flux and the sensible heat flux near the critical Richardson number in detail. In the very stable regime fluxes decrease monotonically with height and a boundary layer resolved by the tower may be very shallow, less than 5 m. Both the stress and the heat decrease rapidly with increasing stability, but the stress falls faster than the heat flux. According to our SHEBA data, the uw-covariance falls as a parabolic function while the heat flux decreases as a linear function. Thus, small but still significant heat flux (several Watts per square meter) and negligibly small stress characterize this situation. In the terms of the transfer coefficients, the Stanton number falls slower than the drag coefficient but faster than the square root of the drag coefficient. As a consequence of the asymmetric decay of the momentum and heat fluxes, the turbulent Prandtl number tends to be less than unity with increasing stability. This result indicates that the heat transfer is more efficient for the very stable regime. This behavior may be associated with the inhomogeneity of the surface temperature and the strong temperature vertical gradient. Small-scale spatial variations of the surface temperature (up to several degrees) generate small-scale advection which enhances temperature variance and the sensible heat flux. According to our SHEBA measurements, the critical bulk Richardson number is about 0.2.