Abstract The computational advances and many times the lack of meteorological measurements in some regions of the world impulse the employment of numerical methods for covering this deficiency of data, and the parameterization is a key factor for use and operation of numerical weather prediction models. The Weather Research & Forecasting (WRF) as one of the most important models has been scarcely parameterized in the Bolivian Altiplano and Titicaca Lake. This work evaluates different physics models and focus the parameterization in the planetary boundary layer (PBL) schemes. This study evaluates the performance of land surface, radiation, convective and microphysics models using WRFv3.8.1. The results showed well-prediction with NOAH, THOMPSON, and RRTMG for land surface, microphysics and radiation schemes, respectively. For the PBL parameterization QNSE, YSU, MYJ, MYNN, and QNSE-FDDA were evaluated for wind speed, temperature, and mixing ratio. The minimum root mean square error were 2.025 m s?1 (QNSE-FDDA) for wind speed, 1.721 K (YSU) for temperature, and 0.786 g kg?1 (MJY) for mixing ratio. In summary, the results have shown that QNSE scheme well-predicts wind speed and water mixing ratio. However, using FDDA only wind speed and water mixing ratio were improved, and the temperature prediction becomes worse. These results suggest an underprediction of incoming solar radiation due to temperature and water mixing ratio were underestimated near the Titicaca Lake.
