Development of a real-time, automated boundary layer height detection algorithm using radar wind profiler data

Boundary layer depth is recognized as one of the key parameters for numerical models to simulate correctly for air quality and surface temperature prediction. Because of this, some of the NOAA/National Weather Service operational forecast models will soon have boundary layer depth as one of the predicted fields available graphically for forecasters to utilize. However, very few measurements of boundary layer depth are routinely available for comparison to the models, and almost never in real-time.

A new technique will be described that relies on 915 MHz radar wind profiler observations of signal-to-noise ratio, vertical velocity, and spectral width to determine boundary layer depth automatically and in real-time. This technique is based on a fuzzy logic algorithm developed previously by Bianco and Wilczak (2002, JTECH, v. 19), but has been expanded to also include information on turbulence intensity as expressed by the radar’s spectral width parameter. Examples of the new algorithm will be shown that demonstrate its ability to detect not only the growing convective boundary layer, but also the collapse of the boundary layer during the evening transition and the collapse of the boundary layer during periods of cloud layer advection aloft. Finally, real-time boundary layer depths will be shown for a radar wind profiler located in New England that will be operating during the time of the conference.