Visualisation of turbulent exchange using a thermal camera

For the purpose of modeling the convective planetary boundary layer (PBL) is usually divided into horizontal layers, where each layer is defined empirically by a set of similarity scales. The latter are needed, in order to get some universal functions from the observed statistics of turbulent exchange. This leads to a framework like the Monin-Obukhov-similarity theory, which is based on the assumption that in the lower PBL a separation of scales is possible and consequently turbulent exchange conditions are determined only by the surface fluxes. This is hard to find in reality and represents only the ideal case, which was known right from the beginning. Large eddies, which scale with PBL height, influence as coherent structures the local exchange processes and question assumptions like stationarity and homogeneity. These coherent structures interact with the flow in the surface layer and can be identified as typical spatial patterns (hair pins or horse shoes, temperature streaks). A trivial finding which one observes when the wind is playing above a wheat field or a water surface. It can be assumed that these active structures impose prominent thermal structures on the surface and their visualisation is the motivation for the work presented here.

A thermal camera (Infratec, Variocam) was mounted at a height of 160 m on the television tower St. Chrischona close to Basel, Switzerland. The camera has a resolution of 320x240 pixels and was looking towards NE. The field of view was a trapezoid with roughly 380/180/420 m (height, 2 parallel sides). The thermal pictures were stored with 0.5 Hz on a laptop. On a bare field in the field of view an energy balance station was installed. There the components of the energy balance equation were measured and with an ultrasonic-anemometer the fluctuations of the wind vector and air temperature were sampled. The measurements took place from 31.7.2007 12h to 2.8.2007 10h.

The time series of the thermal pictures (30min) were decomposed, by subtracting the average picture (mean of 30min) and the trend (deviation of the picture average from the total average). The remaining time series of pictures represent the spatial-temporal fluctuations of surface temperature which were imposed mainly by turbulent exchange. Animations of these time series are shown and discussed with respect to the turbulence measurements in the field.