| Monin–Obukhov Similarity Theory (MOST) is the ubiquitous framework for the estimation
of surface fluxes in atmospheric models. Motivated by tower-measurements, it is conceptually
designed in an ensemble or time averaging context and does not take into account non-local
effects. Hence, the limit of an infinite homogeneous surface is implicit in the assumptions
underlying MOST. This limit is rarely encountered in the coupled land–atmosphere system
which is characterized by broad scale heterogeneities in all quantities. Despite these known
conceptual and practical deficiencies over heterogeneous surfaces, MOST or versions
thereof, such as the MOSAIC approach, are routinely used over heterogeneous
surfaces. In this talk, I present a systematic assessment of the scale margins and
quantify the temporal and spatial limits of the validity of MOST by use of direct
numerical simulation (DNS) of Ekman flow. This investigation of the scale limits of
MOST over a homogeneous surface provides a physically-based temporal and spatial
scale margin below which the interaction of the atmosphere and surface is to be
modelled by physically different approaches. Further, the availability of a DNS dataset
systematically investigating MOST allows an improved representation of interfacial fluxes. |