Earth and Atmospheric Sciences, Department of


Date of this Version



© 2018 American Meteorological Society.

DOI: 10.1175/JTECH-D-18-0019.1


The objective of the research presented is to assess the impact of sensor response and aircraft airspeed on the accuracy of in situ observations collected by small unmanned aircraft systems profiling the convective boundary layer or transecting airmass boundaries. Estimates are made using simulated aircraft flown within large-eddy simulations. Both instantaneous errors (differences between observed temperature, which include the effects of sensor response and airspeed, and actual temperature) and errors in representation (differences between serial observations and representative snapshots of the atmospheric state) are considered. Synthetic data are retrieved assuming a well-aspirated first-order sensor mounted on rotary-wing aircraft operated as profilers in a simulated CBL and fixed-wing aircraft operated through transects across a simulated airmass boundary. Instantaneous errors are found to scale directly with sensor response time and airspeed for both CBL and airmass boundary experiments. Maximum errors tend to be larger for airmass boundary transects compared to the CBL profiles. Instantaneous errors for rotary-wing aircraft profiles in the CBL simulated for this work are attributable to the background lapse rate and not to turbulent temperature perturbations. For airmass boundary flights, representation accuracy is found to degrade with decreasing airspeed. This signal is most pronounced for flights that encounter the density current wake. When representation errors also include instantaneous errors resulting from sensor response, instantaneous errors are found to be dominant for flights that remain below the turbulent wake. However, for flights that encounter the wake, sensor response times generally need to exceed ~5 s before instantaneous errors become larger than errors in representation.