A Big Chill: Examining an Extremely Sharp Cold Front from Three Perspectives

Abstract On 21–22 December 2022, an extremely strong cold front moved southward through the High Plains of the United States. The frontal passage was associated with rapid temperature decreases that rivaled or exceeded the largest that have been reported in the literature. In many locations in Wyoming and Colorado, the temperature dropped from above freezing to below zero Fahrenheit (below −17.8°C) in a few hours. High-resolution observations from an eddy covariance tower in Akron, Colorado, revealed a 2-m air temperature drop of 13.8°C in 10 s and 15.4°C in 30 s, showing that this front effectively represented a discontinuity in temperature. Operational forecast models indicated the potential for a strong cold front up to 9 days in advance but displayed considerable run-to-run variability and uncertainty in the timing of frontal passage. Forecasts from data-driven weather prediction models (trained using deep learning) were generally more accurate than those from physics-based models in terms of the timing and location of the front but did not represent the extreme temperature gradients that occurred. Short-term operational model forecasts provided a detailed depiction of the front but predicted the frontal passage to be slightly slower and weaker than was observed. A high-resolution numerical simulation reveals a frontal structure akin to a density current with a hydraulic head, consistent with past observations of intense cold fronts to the east of the Rocky Mountains. In total, this study provides climatological context and a detailed picture of forecast performance and physical processes for one of the sharpest cold fronts in U.S. records. Significance Statement The High Plains region of the United States routinely experiences the passage of strong cold fronts in the fall, winter, and spring. Yet the cold front on 21–22 December 2022 was among the strongest that has ever been observed in the region. High-resolution observations showed extremely rapid drops in temperature when the front passed: over 15°C in 30 s. Forecast models based on machine learning provided consistently better predictions of the timing of the frontal passage than traditional weather prediction models, whereas the traditional models better represented the sharpness of the front. We also simulated the front with a computer model, which showed that it had a raised “head” at its leading edge, similar to some other intense fronts that were studied previously.