Nocturnal boundary layer controls on minimum temperature variability

Minimum nocturnal temperature is a very relevant variable to a number of applications. At the same time, it is one of the most difficult to be predicted by weather forecast models. Much of this difficulty is associated with the incomplete knowledge of the nocturnal boundary layer dynamics. Temperature minimum predictability decreases in cold nights with weak winds, a situation that favors the disconnection between the surface and upper boundary layer. In this case, therefore, local processes drive scalar variability but, at the same time, very subtle turbulent mixing events may be able to induce appreciable surface temperature changes. Two aspects intrinsic to the nocturnal boundary layer dynamics have, therefore, a crucial role on minimum temperature determination. The first is knowing at which external forcing conditions the surface connects to the upper boundary layer. The second is to understand the intensity and frequency of the intermittent mixing events when the surface is at the decoupled state.

In the present study, the two aspects are analyzed, from both an observational and a modelling perspective. Minimum temperature dependence on mean winds is shown for a network of 40 climate stations in southern Brazil. Many interesting aspects regarding the controls exerted by the nocturnal boundary layer processes on this variable can be inferred from this simple dataset. Among them are an abrupt change in minimum temperatures when the mean winds exceed a threshold, related to the surface coupling state transition, and the fact that in many cases the minimum temperatures do not occur at the lowest wind cases. The minimum temperature forecasts by a mesoscale numerical model are also analyzed for the same stations, and a number of limitations of the model become evident. The same results are, then, analyzed in more detail using turbulence data from micrometeorological towers and single-column schemes for the interaction between the surface and the atmosphere. It is shown that such schemes can reproduce the coupling transition between the surface and the upper boundary layer, but that they still lack the power to capture the intermittent mixing events that occur at the decoupled state. The consequences of this limitation are quantified in terms of the errors introduced to the minimum temperature forecasts.