An Automated Bright-band Height Detection Algorithm Using Wind Profilers

Contact: Paul Neiman

Because knowledge of the melting level is critical to river forecasters and other users, ETL scientists have developed an objective algorithm to detect the bright-band height from profiles of radar reflectivity and Doppler vertical velocity collected with a Doppler wind profiling radar. The importance of melting level information in hydrological prediction is illustrated using the NWS operational river forecast model applied to mountainous watersheds in California (see Figure 1 on the website given). It is shown that a 2000 ft increase in the melting level can triple runoff during a modest 24-hr rainfall event.

The algorithm uses vertical profiles to detect the bottom portion of the bright band, where vertical gradients of radar reflectivity and Doppler vertical velocity are negatively correlated. A search is then performed to find the peak radar reflectivity above this feature, and the bright-band height is assigned to the altitude of the peak. Reflectivity profiles from the off-vertical beams produced when the radar is in the Doppler beam swinging mode provide additional bright-band measurements. A consensus test is applied to sub-hourly values to produce a quality-controlled, hourly-averaged bright-band height. An example of the graphical display used to relay algorithm results to the public in near real time via the internet is shown in Figure 2.

A comparison of radar-deduced bright-band heights with melting levels derived from temperature profiles measured with rawinsondes launched from the same radar site during the Pacific Land-falling Jets Experiment (PACJET) shows that the bright-band height is, on average, 192 m lower than the melting level. The bright-band height is a better estimate of the snow level than the melting level because of the time required for ice particles to melt as they descend. The ability to monitor the bright-band height is likely, therefore, to aid in snow-level forecasting and verification that would prove invaluable to a wide array of user groups, including weather forecasters and hydrologists, emergency managers, and the transportation and ski industries.

Fig. 1. River forecast model simulations of the sensitivity of runoff to changes in melting level for four river basins in California. The simulations were conducted by the California/Nevada River Forecast Center using the National Weather Service River Forecast System (NWRFS). Each run used a different melting level ranging from low to high elevations within the basin. Initially, each basin was brought to a mid-winter soil moisture condition by adjusting the parameters of the Sacramento Soil Moisture Accounting Model of the NWRFS. The 24-hour quantitative precipitation forecast used to drive the model is shown in the upper left. The peak streamflow for each run is plotted. The posted numbers give the approximate percentage of basin area below the altitude corresponding to the melting level. These percentages were determined by linearly interpolating the area elevation curves generated for each basin in the SNOW-17 module of the NWRFS.

Fig. 2. (a) Example of a bright-band height (BBH) image from the Pacific Land-falling Jets Experiment displayed on a real-time data web page (http:// www7.etl.noaa.gov/data/). The colored rectangles display hourly averaged values of Doppler vertical velocity (DVV; positive downward) recorded by the wind profiler at Bodega Bay, California (BBY, elev. 12 m) on 24 February, 2001. The BBH data are indicated by black dots. The time axis proceeds from right to left. Heights are above mean sea level (MSL). The image has been annotated to indicate patterns of DVV corresponding to radar backscatter from different atmospheric media. Noise refers to measurement noise caused by low signal power. (b) Time-height cross section of wind barbs (flag = 25 m s-1; full barb = 5 m s-1; half barb = 2.5 m s-1) and isotachs (m s-1) of the zonal wind component recorded by the wind profiler at Bodega Bay for the same period as in (a). A descending warm front is dilineated by a region of enhanced speed and directional wind shear. The black dots are the same BBHs plotted in (a).

Related Links:

• Regional Weather and Climate Applications Division
• Pacific Landfalling Jets Experiment