The sandstorm, which is a common natural disaster, is mechanically characterized by a particle-laden flow experiencing wall turbulence. This work investigates a real sandstorm that was measured at the Qingtu Lake Observation Array through a lens of wall-turbulent flow dynamics. A non-stationary signal processing method is proposed based on the time-varying mean and adaptive segmented stationary method, and the evolution of turbulent kinetic energy during the entire sandstorm process is revealed.