Abstract
The variability of trace gases at surface monitoring stations is affected by multi-scale meteorological processes that are particularly complicated over mountainous terrain. A detailed understanding of the impact of meteorological processes on mountaintop trace gas variability is vital to help distinguish trace gas measurements affected by local pollutant sources from measurements representative of background mixing ratios. Although this knowledge is present for tall mountaintops, the trace gas variability at low mountaintops is more complicated because of their location at the transition between the planetary boundary layer (PBL) and free atmosphere (FA). The goal of this dissertation is to improve scientific understanding of the physical processes affecting trace gas mixing ratios sampled at low mountaintops. To fulfill this goal, this dissertation presents the first use of trace gas measurements from Pinnacles (38.61 N, 78.35 W, 1017 m above sea level [msl]), a mountaintop monitoring site in the Appalachian Mountains. This dissertation also uses in situ meteorological measurements from the site, trace gas and meteorological observations from nearby stations, and numerical simulations. Observations from Pinnacles indicate that wind shifts are an important driver of the diurnal trace gas variability. Cold fronts induce increases in CO and CO2 mixing ratios of >100 ppb and >20 ppm, respectively, on timescales <3 h. Wind shifts from the northwest to the south on fair weather days yield daytime CO increases. On days without these wind shifts, CO mixing ratios decrease. On this subset of days, the height of the afternoon valley PBL height, z_i, relative to the ridgetop was expected to be a significant driver of the trace gas variability. To investigate this hypothesis, z_i estimates were needed for the Page Valley upwind of Pinnacles. To this end, a technique was developed to estimate z_i over the Page Valley using observations from Dulles Airport, IAD (38.98 N, 77.49 W, 87 m msl), which is the sounding station nearest Pinnacles, as well as output from reanalysis products and numerical simulations with the Weather Research and Forecasting model. Page Valley z_i were found to be 200-400 m higher than those from IAD, and thus an offset was applied to better estimate z_i over the Page Valley from the IAD soundings. The corrected z_i were then used to investigate the influence of PBL dilution on the trace gas measurements from Pinnacles. Days with z_i below the ridgetop height have a daytime CO increase caused by the transport of valley PBL air to the mountaintop. Days with z_i exceeding the ridgetop exhibit a daytime CO decrease caused by dilution and entrainment of FA air. Case studies and numerical simulations with a Lagrangian Particle Dispersion Model (LPDM) were used to understand these physical processes in more detail and investigate how mountaintop measurements can best be used in applications requiring regionally-representative values. On days when z_i exceeds the ridgetop, afternoon trace gas measurements from low mountaintops are most similar to measurements from the tops of tall towers in flat terrain, based on comparisons between LPDM simulations with topography included and LPDM simulations from which the topography was removed. In summary, in this dissertation it is found that mesoscale to synoptic scale wind shifts are important drivers of the trace gas variability at low mountaintops and result in trace gas characteristics typical of other, much taller mountaintops. The wind shifts result in different upwind emission sources being sampled and can help estimate emissions and reduce flux estimate uncertainties over upwind areas. On days with constant winds and PBLs exceeding the ridgetop height, trace gas measurements have characteristics of monitoring sites in flat terrain and can be used in applications requiring regionally-representative values.
