Observational and numerical studies of rapid warm front passages

As opposed to cold fronts, few studies have been conducted on warm fronts since the development of the Norwegian Wave Cyclone model. Moreover, no systematic study has clarified just how surface warm fronts move poleward, since the direct ageostrophic circulation perpendicular to the front is always directed toward the warm air, and the warm air flowing poleward at the warm front should move over the denser, more stable cool layer ahead of the front. Also, not all warm fronts are weak, with the temperature changes associated with some matching or exceeding that of the strongest of cold fronts. Thus, the causes of rapid warm front passage and the general mechanisms behind warm frontal movement are poorly understood and are investigated in this study.

One such warm front passed through Oklahoma during the nighttime hours on 12 January 2005. The Oklahoma Mesonet provides a detailed set of data for this event. The front increased the temperature by more than 10°C in less than 15 minutes at multiple locations in eastern Oklahoma. To the west, central Oklahoma showed steady temperature increases common to typical warm fronts. We hypothesize that rapid warm frontal movement is a result of intense mixing near the frontal zone. Since this event occurred during the night and in the absence of any precipitation, the intense mixing is likely a direct result of mechanical turbulence. This is supported by the overnight development of a low level jet and a coincident 12 hour increase in wind speed of 41 knots over a depth of 250m near the surface. The depth of the cold air that must be mixed is obtainable via the hydrostatic equation and detailed five minute mesonet observations.

Investigation of short-range forecasts by both the RUC and ETA operational models showed that neither model was able to duplicate the sudden increases associated with the Jan. 12 warm front. In an attempt to recreate the intense nature of the front, experiments with the WRF-ARW numerical model are performed for this event. We investigate the response of the WRF forecasts to variations in horizontal and vertical resolution, and in the specification of the various boundary layer schemes. The results will reveal what model configurations are required to accurately simulate intense mixing processes associated with frontal passages.