Atmospheric boundary-layer turbulence induced surface temperature fluctuations. Implications for TIR remote sensing measurements

Atmospheric turbulence in both surface (SBL) and planetary (PBL) boundary-layers govern rapid temporal fluctuations of the surface temperature (Ts), with potentially important resulting errors on instantaneous satellite measurements in the thermal infrared (TIR). An experiment designed to evaluate the error, and its reduction with spatial resolution (expected from ergodicity properties), of Ts is described. It is based on the analysis of time series of Ts at different resolutions reconstructed from sequences of images acquired at high frequency using a helicopter borne TIR camera over 5 different surfaces in the south-west of France, in the 2010 summer. Unfortunately the non-respect of a stationary flight resulted in a contamination of the data by directional effects; the filtering process developed to correct for this also eliminated the signature of low frequency PBL turbulence related to Ts measurements, and made only possible the study of high frequency turbulence impact on Ts. For decreasing resolution (i.e., increase in pixel size), the contribution of the SBL turbulence to error rapidly decreases with resolution (i.e., pixel size increase); it reaches ± 0.5 °C at 50 m, the resolution selected for recent spatial mission projects, with only little variations thereafter. Recommendations are made for future experiments to evaluate the overall error induced by atmospheric turbulence on satellite Ts measurements.