Quantifying the Wind-Induced Bias of Rainfall Measurements for the Thies CLIMA Optical Disdrometer

Abstract

The wind-induced bias of rainfall measurements obtained from non-catching instruments is addressed in this work with reference to the Laser Precipitation Monitor (LPM) optical disdrometer manufactured by Thies CLIMA. A numerical simulation approach is adopted to quantify the expected bias, involving three different models with increasing complexity. Computational Fluid-Dynamics simulation of the airflow field around the instrument with an embedded Lagrangian particle-tracking module to obtain raindrop trajectories are performed by solving the Unsteady Reynolds Averaged Navier-Stokes (URANS) equations and a Large Eddy Simulation (LES) model. URANS-uncoupled, LES-uncoupled, and LES-coupled approaches are tested to assess the impact of modeling the airflow turbulent fluctuations in detail. Due to the non-radially symmetric external shape of the instrument, various combinations of the wind speed and direction are considered. Catch ratios for monodisperse rain are obtained as a function of the particle Reynolds number and the wind direction and fitted to obtain site-independent curves to support application of the simulation results. Based on literature expressions to link the drop size distribution of real rainfall events with the rainfall intensity (which instead depend on the local rainfall climatology at the measurement site), sample collection efficiency curves are obtained from the catch ratios of monodisperse rain. The resulting adjustment curves allow rainfall measurements to be corrected using either a real-time or post-processing approach. However, at high wind speed and assuming that the wind blows parallel to the instrument sensing area, the instrument may fail to report precipitation altogether.