Near-surface atmospheric conditions, especially the temperature and winds, are characterized by their diurnal variations. Accurate representation and forecast of the diurnal variations are the essential components of numerical modeling and weather prediction. However, it is commonly challenging to accurately simulate and predict diurnal variations of near-surface atmospheric conditions over complex terrain, especially over the mountainous areas.

In this study we evaluate the diurnal variation of near-surface temperature and winds from the numerical simulations generated by mesoscale community Weather Research and Forecasting (WRF) model. The model simulated surface temperature at 2-meter height and winds at 10-meter height are compared with these observations from surface mesonet observations in several different weather scenarios (winter inversion, cold front and low-level jet etc.) over the Inter-mountain West of US. Preliminary results show large discrepancies between model generated diurnal variations and observations in some cases. The mechanism and causes of these differences are further investigated. The impact of errors in the model simulated diurnal variations on assimilation of surface observations are also examined.