During the Verification of the Origins of Rotation in Tornadoes EXperiment-Southeast project (VORTEX USA), held in the Southeast United States, many in-situ and ground-based remote sensing platforms were deployed to monitor the dynamic and thermodynamic state of the atmosphere to explore the storms and conditions that make tornadoes especially dangerous in the southeastern United States. In this study, data collected at one of the VORTEX USA sites (Courtland, AL) by a 449-MHz radar wind profiler (RWP) with associated radio acoustic sounding system (RASS), a ceilometer, and various surface observations are used to assess the operational High Resolution Rapid Refresh (HRRR – 3 km horizontal grid spacing) numerical weather prediction model. The RWP and RASS datasets cover the period from the end of 2019 to present, with additional information on cloud-base height available during the last year and a half from the ceilometer. The HRRR model errors of surface variables and wind speed, direction, and virtual temperature within the boundary layer are evaluated as a function of the year, season, daily cycle, and forecast horizon over a multi-year period. Moreover, taking advantage of the additional information on cloud coverage observed by the ceilometer over the last year and a half, the analysis digs deeper into the representation of the HRRR model errors as a function of cloud characterization.