Modeling and Lidar Study on Ozone Over the Chesapeake Bay During OWLETS-2

Abstract

This study focuses on the distribution of ozone (O3) concentration near the Chesapeake Bay, USA (hereafter CB) by integrating observations and model simulations. The motivation of this work is to understand reasons causing the horizontal and vertical distribution of pollutants (mainly O3) near the CB. The O3 exceedance over the CB happens very frequently during summer and the Maryland Department of Environment intends to find out the reasons in order to make policy-related decision. The observation data used in this study are from the Ozone Water-Land Environmental Transition Study-2 (OWLETS-2) field campaign, including observations from O3 lidar, Doppler wind lidar, ozonesonde. The mesoscale model employed is Weather Research and Forecasting model coupled with chemistry (WRF-Chem) version 3.9.1. The anthropogenic emission dataset is from National Emission Inventory 2011 (NEI-2011), including various emission species, e.g., CO, NOX, SO2, NH3, PM2.5, PM10, etc. The meteorological initial and boundary conditions are from the Northern American Regional Reanalysis (NARR) dataset, which is a high-resolution combined model and assimilated dataset from the National Centers for Environmental Prediction (NCEP). There are several findings of this study based on the model simulations and ground-based observations. Actually, at the beginning of study, we considered two different versions of anthropogenic emissions from NEI-2005 and NEI-2011 developed by the Environment Protection Agency (EPA). EPA added the anthropogenic emissions over CB from boats and ships while updating from NEI-2005 to NEI-2011. For model performance evaluation, we employed AirNow surface hourly O3 mixing ratio diurnal variation and compared it with model simulations. For instance, at Essex site near Baltimore City, observed O3 has a strong diurnal variation, with minimum (25 ppbv) just after sunrise (05:00 EST), and with maximum (75 ppbv) around afternoon (15:00 EST). Even the model simulation has a good agreement with the observation, it underestimates the mean O3 mixing ratio by about 15-20 ppbv. Both the surface and 700 mb level horizontal spatial distribution of O3 indicate the higher O3 concentration over the north-middle CB, with surface O3 mixing ratio of 40-50 ppbv and 700 mb level O3 mixing ratio of 60 ppbv, which means the surface O3 was lifted up after production. The vertical profiles of wind of both model and Doppler wind lidar match very well, indicating that the model captured the vertical variation of wind. However, the vertical profiles of O3 from model simulation, ozonesonde, and O3 lidar suggests that model simulation underestimated the O3 from surface to 4.5 km. In addition, the model simulation captured the vertical mixing of O3 from surface to 2 km, while misses the O3 variation above 2 km. In order to study the influence of bay breeze on the O3 small scale transport, three vertical cross sections through the CB from west to east at the northern, middle, and southern CB. The results show that higher O3 concentration above the CB. The bay breeze over the southern CB is stronger than the northern CB. The planetary boundary layer height over the CB is dramatically lower than the surrounding land in the day, which contributes to the surface higher O3 concentration over the CB.