Application of active optical sensors to probe the vertical structure of the urban boundary layer and assess anomalies in air quality model PM2.5 forecasts

In this paper, the simulations of the Weather Research and Forecast (WRF) and Community Multiscale Air Quality (CMAQ) Models applied to the New York City (NYC) area are assessed with the aid of vertical profiling and column integrated remote sensing measurements. First, we find that when turbulent mixing processes are dominant, the WRF-derived planetary boundary layer (PBL) height exhibits a strong linear correlation (R > 0.85) with lidar-derived PBL height. In these comparisons, we estimate the PBL height from the lidar measurements using a Wavelet Covariance Transform (WCT) approach that is modified to better isolate the convective layer from the residual layer (RL). Furthermore, the WRF-Lidar PBL height comparisons are made using different PBL parameterization schemes, including the Asymmetric Convective Model-version2 (ACM2) and the Modified Blackadar (BLK) scheme (which are both runs using hindcast data), as well as the Mellor-Yamada-Janjic (MYJ) scheme run in forecast mode. Our findings show that the correlations for these runs are high (>0.8), but the hindcast runs exhibit smaller overall dispersion (≈0.1) than the forecast runs. We also apply continuous 24-hour/7-day vertical ceilometer measurements to assess WRF-CMAQ model forecasts of surface PM2.5 (particulate matter has aerodynamic diameter <2.5 μm). Strong overestimations in the surface PM2.5 mass that are observed in the summer prior to sunrise are particularly shown to be strongly connected to underestimations of the PBL height and less to enhanced emissions. This interpretation is consistent with observations that TEOM (Tapered Element Oscillating MicroBalance) PM2.5 measurements are better correlated to path-integrated CMAQ PM2.5 than the near-surface measurements during these periods.