Spatial and temporal variability of surface temperature of a forest canopy

Acquiring the surface temperature is important in various research fields needing to quantify the surface energy balance (e.g. meteorology, hydrology, climatology, etc.). For such task, high-spatial resolution satellite thermal infrared (TIR) measurements are promising, but due to constraints in the specifications of space missions, TIR satellite images are prone to temporal and spatial variability in the measurement acquisition in relation with the rapid feedback between the surface temperature and atmospheric turbulence. This variability in measurement acquisition may limit the data accuracy. During daytime, the fluctuations of surface temperature are governed by large-scale convective eddy motions in the outer atmosphere and local coherent eddies forming over vegetation canopies. To estimate the possible error related to these fluctuations, it is necessary to understand the extent of their length- and time-scales.

With such motivation, an experiment was conducted in 2015 in the Landes forest in France. A TIR camera overlooking a maritime pine forest stand (15-20 m high) was set up to acquire the surface temperatures at high spatial and temporal resolutions (≈10 m, 10 Hz), together with wind velocity from a sonic anemometer. A scaling analysis of the measurements (Fourier, wavelets, etc.) is performed to identify the scaling regime of the surface temperature fluctuations following the atmospheric stability. We discuss these results in the context of satellite thermal imagery uncertainties.