Statistical comparison between Hysplit sounding and lidar observation of planetary boundary layer characteristics over New York City

Abstract

The need to characterize in a robust way Planetary Boundary Layer (PBL) heights is crucial as in air quality forecast and transport models. In particular, incorrect determination of PBL heights can severely distort the surface air quality predictions such as PM2.5. Local properties and morphological features can influence PBL dynamics through local circulation phenomena such as the sea-breeze development as well as influences from the Urban Heat Island Canopy resulting in multiple layers that need to be resolved. In this paper, based on a combination of wavelet and image processing methods, we develop methods to quantify multilayer PBL's and assess their dynamics with meteorological measurements including temperature, wind and humidity profiles. In particular, meteorologically based PBL heights based on both the Potential Temperature Gradient and Richardson Number are compared against both lidar and ceilometer measurements. It is shown that in general, the Potential Temperature Gradient method is better correlated to the PBL dynamics. Meanwhile, the Hysplit model provides sounding data which can be used for comparison between actual sounding and lidar measurements. On the other hand, when strong atmospheric instability is present or layering develops, the comparison between different methods can provide information about the PBL internal structure. Further comparisons with air quality models such as MM5 are also made and illustrate the difficulty in these models properly predicting the PBL dynamics seen in urban areas.