Experimental evidence and LES suggest that the boundary-layer height is not the only length scale in the CBL. Jonker et al. (H.J.J Jonker, P.G. Duynkerke and J.W.M. Cuijpers, J. Atmos. Sci.,56, 801 - 808, 1999) reported steadily growing length scales for scalars based on large eddy simulations: after about 40 eddy turn-over time scales the dominant length scale of mid-CBL horizontal scalar fields was three times the boundary layer depth and was still increasing, except for scalar entrainment ratios close to -0.5. At the same time, the dominant length scales for velocity fields remained stationary (of the order of the boundary-layer depth), while those for temperature appeared to be an intermediate case.

In 2000 a laboratory experiment was carried out at CSIRO Atmospheric Research in Melbourne, Australia. Notwithstanding some experimental difficulties, the laboratory results supported the above LES findings for horizontal scalar fields. We shall briefly describe the experiment and present the major results. In addition we shall give some theoretical arguments to explain this behaviour.