Investigating Taylor’s frozen turbulence hypothesis in the surface layer at an ideal desert field site using fibre-optic distributed temperature sensing

Taylor’s frozen turbulence hypothesis is the most critical assumption through which time-resolving sensors may be used to derive statistics of the turbulent spatial field. Namely, it relates temporal autocorrelation to spatial correlation via the mean wind speed and is invoked in almost all boundary layer field work. Nevertheless, the conditions and scales over which Taylor’s hypothesis is valid remain poorly understood in the atmospheric boundary layer.

As part of the Namib Turbulence Experiment (NamTEX) campaign in March 2020, a pseudo-3D fibre-optic distributed temperature sensing (DTS) array was installed within a 300 x 300 m area in the Namib desert. The array is X-shaped in plan view and contains 16 measurement heights from 0.45 m to 2.85 m. Fibre-optic sensing provides air temperature measurements at unprecedented spatio-temporal density (0.25 m horizontally, 0.17 m vertically, and 1 Hz) and was coupled with a vertical array of traditional sonic anemometer point measurements to investigate the relationship between spatial and temporal temperature fields. The Namib provides an ideal location for fundamental boundary layer research: homogenous flat surfaces, no vegetation, little moisture, strong solar forcing, regular and repeated clear-sky conditions, and a wide range of atmospheric stabilities.

Using the NamTEX DTS array we present the first field investigation of Taylor’s hypothesis that considers boundary layer stability and is independent of wind direction. A novel method of 2d horizontal cross-correlation between all possible points of a single height of the DTS is employed to produce spatial ‘maps’ of the turbulent flow, whose velocity, direction, and size may be tracked through time.