Sensitivity of Numerical Simulations of an Ephemeral Shallow Cold Fog to Snow Cover in Initial Conditions during the CFACT Field Campaign

Abstract Accurate predictions of mountainous cold fog are crucial but challenging because poor forecasts often arise from imprecise surface initial conditions. This study evaluates how initial conditions from five U.S. weather analysis products influence the Weather Research and Forecasting (WRF) Model at gray-zone resolution (∼333 m) when simulating an ephemeral shallow cold fog event on 19 February 2022 in Heber Valley, northern Utah, during intensive observing period (IOP) 8 of the Cold Fog Among Complex Terrain (CFACT) campaign. Results show that all WRF simulations capture large-scale weather patterns well. However, WRF initial conditions derived from coarse-resolution datasets (>3 km) overestimate snow cover in Heber Valley, even though only trace snow was observed during CFACT IOP 8. This erroneous snow cover leads to excessive daytime shortwave radiation reflection, notable nighttime near-surface cold biases, and excessive saturation. These conditions promote the erroneous moisture release and persistent optically thick freezing fog, contradicting observations. In contrast, simulations using high-resolution analysis products (≤3 km) show minimal biases in shortwave and longwave radiation, near-surface temperature, dewpoint, and mixing ratio—linked to better representation of observed trace-snow-cover areas, as confirmed by sensitivity tests. This study highlights the importance of adopting high-resolution snow cover to improve cold fog simulations over complex mountainous terrain. The results also underscore the critical radiative effects of snow cover on near-surface atmospheric conditions and the consequent influence on the formation and evolution of cold fog. Furthermore, the numerical results confirm that high near-surface vertical resolution is necessary for better prediction of the near-surface temperature inversions and ephemeral fog events. Significance Statement Accurate prediction of cold fog in mountainous regions is vital for transportation safety but remains difficult, partly due to uncertainties in snow representation. We evaluated the influence of snow initial conditions from five U.S. weather analysis products on simulations of an ephemeral shallow cold fog event in Heber Valley, Utah, observed during the Cold Fog Among Complex Terrain (CFACT) field campaign. Our findings highlight that adopting high-resolution snow analysis and fine near-surface vertical resolution is critical for reliably simulating and forecasting mountainous cold fog. Our results also imply that snow cover can promote the formation of cold fog by reflecting shortwave radiation and lowering near-surface temperatures.