KOUN Case Study Detecting the melting layer in a winter storm Author: Kevin Scharfenberg, CIMMS/OU Last Updated: 15 January 2003

Introduction

On 3-4 December 2002, a winter storm moved across Oklahoma. The storm was responsible for several inches of snow across northwest and north central Oklahoma, a band of freezing rain west and north of Oklahoma City, and rain elsewhere. The hydrometeor classification algorithm output from this event, and a verifying sounding, are studied.

Indentifing the melting layer

Although a "bright band" of enhanced reflectivity typically marks the melting layer in cold season precipitation events, the reflectivity image (Figure 1) does not clearly show a "bright band". The phase change between rain and snow is much more clearly evident in the differential reflectivity image (Figure 2) where differential reflectivity increases suddenly with range by about 1 dB. This increase is often observed in the melting layer, and is likely caused by aggregation. That is, the falling snow crystals become wet and begin to stick to each other. The correlation coefficient image (Figure 3) shows that while values are above 0.99 in the rain, values range from 0.9 to 0.95 in the mixed phase regime.

Figure 1: Reflectivity

Figure 2: Differential Reflectivity

Figure 3: Correlation Coefficient

When considering the increasing altitude of the radar beam with range, the hydrometeor classification algorithm (HCA) attempts to mark the level above which mostly snow exists and below which mostly rain exists. For the radar scan corresponding to Figures 1 through 3, the height of that level detected by the HCA (Figure 4) ranges from 1480 meters above radar level to the northwest of the radar to 1850 m above radar level to the southwest. This is consistent with the range where differential reflectivity increased in Figure 2.


Figure 4: Hydrometeor classification algorithm output.

Sounding verification

A special sounding was released at the Norman, OK office of the National Weather Service at the time of the radar images in Figures 1-4. This sounding was used to verify the actual melting level altitude. The sounding (Figure 5) found the melting level to be about 2350 m ARL. At first glance, this may appear inconsistent with the KOUN measurement of the rain-snow line between 1480 and 1850 m ARL. However, note the sounding indicates the layer of air just below the melting level is subsaturated. Observational studies indicate snow falling below the melting level into subsaturated area may fall more than 500 meters before melting. Therefore, the KOUN measurement is reasonable, and is also consistent with reports in northwestern Oklahoma of snow falling with surface temperatures of 2° C to 3° C.

Also note in the HCA image (Figure 4) that the detected melting level is higher to the south (1850 m ARL) than to the north (1480 m ARL). This is consistent with the weather situation on 3 December 2002, when the cold air mass was cooling and deepening from the north.


Figure 5: Norman, OK sounding corresponding to the radar image in Figure 4.

Conclusions

Since a bright band was difficult to identify, the polarimetric products were valuable in determining the location of the melting layer in this case. The differential reflectivity field could be used to mark the location of ice aggregates, which was also marked by a decrease in the correlation coefficient. The hydrometeor classification algorithm correctly indicated a lower altitude of melting to the northwest of the radar than to the south, associated with the deeper and colder air mass to the northwest. While the sounding suggested the freezing level was at least 500 m higher, the KOUN observation is consistent with previous observations of a deep layer of melting below the melting level in subsaturated environments.

Related Links

• National Severe Storms Laboratory (NSSL)
• JPOLE: Joint POLarimetric Experiment
• NSSL Polarimetric Radar Page

Acknowledgements

This case study would not have been possible without the dedicated work of the scientists who maintain, operate, and collect data from KOUN radar. Thank you!