|dc.description.abstract||Nitrogen oxides (NOx = NO + NO2) play a crucial role in the formation of ozone and has significant impacts on the production of secondary organic and inorganic aerosols, thus affecting human health, global radiation budget, and climate. Accurate knowledge of NOx emissions is essential for relevant scientific research and air pollution control policies. This thesis evaluates current estimates of anthropogenic and natural NOx emissions over the United States and improves model’s prediction of surface ozone concentrations by using a 3-D Regional chEmistry and trAnsport Model (REAM) and various types of observations and investigate the impact of thunderstorms on surface NOx and O3 concentrations.
The diurnal cycle of NO2 is a function of emissions, advection, deposition, vertical mixing, and chemistry. Its observations, therefore, provide useful constraints in our understanding of these factors. The REAM simulated diurnal cycles are evaluated by using the DISCOVER-AQ campaign measurements, EPA Air Quality System (AQS) observations, and OMI and GOME-2A tropospheric vertical column densities (TVCDs) products in July 2011 over the Baltimore-Washington region. The model simulations are in reasonably good agreement with the observations except that PANDORA measured NO2 TVCD show much less variation in the early morning and late afternoon than simulated in the model. High resolution (4 km in the horizontal) model simulations are also performed to examine the effects of emission distributions. The overestimation of NO2 concentrations from the 4-km REAM simulation in contrast to the well reproduction of observations by the 36-km REAM suggests that the 2011 National Emission Inventory (NEI2011) provide a good estimate of NOx emissions at the 36-km scale but can’t resolve NOx emission distributions at the 4-km resolution. By analyzing model simulations with the observations, the thesis shows that the diurnal emission profile of NOx is different over the weekend from the weekdays and that weekend emissions are about 1/3 lower than weekdays. Observed ozone concentrations can be used to evaluate NOx and volatile organic compound (VOC) emissions by using their relationships with ozone concentrations. The thesis shows that the time when ozone reaches its daily maximum (peak time) is also related to NOx and VOC emissions. Through model sensitivity analyses of REAM in July 2011 over the contiguous United States (CONUS), it is found that ozone peak values are more sensitive to NOx emissions while ozone peak time is more sensitive to VOC emissions in the eastern United States. By such relationships and the comparison between observations and model results, we find that the underestimation of soil NOx emissions leads to a low bias of simulated ozone peak value in the South, while the overestimation of biogenic isoprene emissions results in earlier than observed ozone peak time in the Central, South and Southeast regions. The simulated formaldehyde columns, which are higher than satellite measurements, confirm the latter. We illustrate the nonlinear relationships among NOx emissions, NO2 TVCDs, and NO2 surface concentrations using the simulations of REAM for July 2011 over the CONUS. The variations of NO2 surface concentrations and TVCDs are generally consistent and reflect well anthropogenic NOx emission variations for high-NOx emission regions. For low-NOx emission regions, however, nonlinearity in the emission-TVCD relationship makes it difficult to use satellite observations to infer anthropogenic NOx emission changes. The analysis is extended to 2003 – 2017. Similar variations of NO2 surface measurements and coincident satellite NO2 TVCDs over urban regions are in sharp contrast to the large variation differences between surface and satellite observations over rural regions. We find a continuous decrease of anthropogenic NOx emissions after 2011 by examining surface and satellite measurements in CONUS urban regions, but the decreasing rate is lower by 9% - 46% than the pre-2011 period. By comparing observed ∆O3 (hourly change of O3 concentrations) and ∆NOx with/without lightning events, we find that generally, thunderstorms decreased ∆O3 in the daytime due to the dominant role of solar radiation reduction reaching the surface and increased ∆O3 during the nighttime due to convective downdrafts and increased nocturnal boundary layer mixing. With our adjustment of downdraft mass fluxes (DMFs) and eddy diffusivity coefficients during nighttime thunderstorm events which are underestimated by the Weather Research Forecast (WRF) model, REAM well reproduces the observed characteristics and produces a bimodal post-convection lightning NOx shape with one peak near the surface. Sensitivity simulations show that lightning NOx contributes 2.4 – 3.6 ppb to MDA8 in the southeast U.S.||