The infrared measurement cascade: Connecting large scale meteorologically induced surface temperature perturbations to local spatial velocity structures

The evolution of micrometeorological measurements has been recently manifested by developments in methodological and analytical techniques using spatial surface brightness temperature captured by infrared cameras (Schumacher et al. 2019, Katurji and Zawar-Reza 2016). The Thermal Image Velocimetry (TIV) method can now produce accurate 2D advection-velocities using high speed (>20Hz) infrared imagery (Inagaki 2013, Schumacher 2019). However, to further develop TIV methods and achieve a novel micrometeorological measurement technique, all scales of motion within the boundary layer need to be captured.

Spatial observations of multi-frequency and multi-scale temperature perturbations are a result from the turbulent interaction of the overlying atmosphere and the surface. However, these surface signatures are connected to the larger scales of the atmospheric boundary layer (McNaughton 2002, Träumner 2015). When longer periods (a few hours to a few days) of spatial surface brightness temperatures are observed, the larger scale information needs to be accounted for to build a comprehensive understanding of surface-atmospheric spatial turbulent interactions. Additionally, the time-frequency decomposition of brightness temperature perturbations shows longer periods of 4-15 minutes superimposed over shorter periods of ~ 4–30 seconds. This suggests that that boundary layer dynamic scales (of longer periods) can influence brightness temperature perturbations on the local turbulent scale. An accurate TIV algorithm needs to account for all scales of motion when analysing the time-space variability of locally observed spatial brightness temperature patterns.

To analyse these propositions temporally high resolved geostationary satellite infrared data from the Himawari 8 satellite was compared to near-surface and high speed (20 Hz) measured air and brightness temperature using thermocouple measurements and infrared cameras. The satellite provides a temporal resolution of 10-minutes and a horizontal resolution of 2 by 2 km per pixel and therefore captures the atmospheric meso γ and micro α scale which signals are usually active for ~10 minutes to < 12 hours. Moreover, the Himawari 8 brightness temperature was used to create the near-surface mean velocity field using TIV. Afterwards, the velocity field was compared to the in-situ measured wind velocity over several days during January 2019.

The results show that the atmospheric forcing from the micro α scale to lower atmospheric scales has a major impact on the near-surface temperature over several minutes. A significant (p-value: 0.02) positive covariance between the Himawari 8 measurement and the local measured temperature 1.5 cm above the ground on a 10 minute average, specifically concerning cooling and heating patterns, has been found.

Further analysis demonstrates that the retrieved near-surface 2-D velocity field calculated from the Himawari 8 brightness temperature perturbations is correctly representing the mean velocity. This finding allows the classification of meso-scale atmospheric forcing and its direct connection to local scale turbulent 2-D velocity measurements. This extends the TIV algorithm by a multi-scale component which allows to address inter-scale boundary layer analysis from a new point of view. In respect to the current findings a new experiment will focus on the repeated induced local velocity patterns from large scale forcing which will be measured through the surface brightness temperature.