Identification and Prediction of Cold-Air Damming in the Northeast United States: A Comparison of Numerical Models

Cold-air damming is a mesoscale phenomenon that occurs when a cold dome is created along the lee side of mountain ranges. Cold air becomes trapped in part because of low-level synoptic flow towards the lee side of mountains, preventing the cold dome from escaping. When this condition occurs, temperatures are cooler on the inside of the cold dome than those outside of the dome. Because of these variations in temperatures, cold season rain showers in the surrounding area can result in freezing rain or ice pellet events in the dome. Despite being a common weather phenomenon, cold-air damming has eluded accurate detection and prediction by most numerical models. Common difficulties the models have include: the timing of the event, the degree to which the trapped air is colder than the surrounding air, the precise location of the damming, and even failure of the model to detect cold air damming at all. Among the explanations for these model difficulties is that the resolution of the model may not be fine enough to resolve the phenomenon. Other limitations may arise due to the type of coordinate system the model uses and the schema for computing dynamical, thermodynamic, and physical processes. Given the recognized limitations of past models, the goal of this research is to determine if more recent versions of commonly used weather models, such as the GFS and the HRRR, can more accurately forecast cold-air damming. An additional goal is to assess and compare these weather models’ observed outcomes and find which has the best predictive accuracy.