Wind Estimates in the Lower Atmosphere from a Vehicle Dynamic Model Applied to a Multi-rotor Copter

Multi-rotor copters are increasingly used in atmospheric boundary layer research. In this poster, we present some results of a recent field campaign that had the major goal to demonstrate the use of unmanned aerial vehicles in probing the lower atmosphere. The field campaign, 2018 ISARRA Flight Week, deployed many different types of unmanned aerial vehicles, as well as conventional instrument platforms, to investigate the morning transition and convection initiation in the San Luis Valley, Colorado. We focus on the estimation of wind using a rigid-body dynamic model of the 3DR-Solo quadcopter. The model developed for the quadrotor is a quasi-linear parameter varying (qLPV) model characterized from system identification flight experiments. This approach expands the flight envelope from which wind estimates can be derived beyond hover flight. We will present results of our wind estimates and compare them with state-of-the-art meteorological instruments installed on a 10-m tower and with Doppler lidars. We also collected temperature and humidity data on our quadcopter and compare the retrieved thermodynamic structure with the estimated wind structure to quantify boundary layer growth and convection potential. We demonstrate the strengths of our profiling approach and provide recommendations for accurate retrieval of wind, temperature, and humidity data from a multi-rotor copter for boundary layer research.