Investigating the spatial representativeness of mountaintop carbon monoxide measurements using aircraft observations and numerical modeling

Knowledge of the spatial representativeness of trace gas measurements is important for many applications, including carbon cycle studies and air quality applications. In the present study, we seek to quantify the spatial representativeness of mountaintop carbon monoxide (CO) measurements from Pinnacles, a new monitoring site in the Blue Ridge Mountains in northern Virginia, from which CO, CO2, and a suite of supporting meteorological measurements have been made since 2008. To determine the representativeness of the CO measurements, we use vertical CO profiles obtained from multiple aircraft flights made 18-29 July 2011 over several nearby locations in northern Virginia and Maryland, and run numerical simulations for this period.

On most days in the study period, we find good agreement between Pinnacles CO measurements (~1000 m above mean sea level [msl]) and CO measured approximately 1000 m msl in the aircraft profiles, indicating that Pinnacles' measurements are largely regionally-representative. On a few days, however, the aircraft CO mixing ratios 1000 m msl are nearly twice as large as the CO mixing ratios measured at Pinnacles, suggesting that at times Pinnacles may not be sampling regional pollution but instead sampling air masses representative of background concentrations. We hypothesize that differences in boundary layer height and stability play an important role in determining the representativeness of the Pinnacles CO measurements in these different scenarios. We investigate this hypothesis by running simulations with the Weather Research and Forecasting (WRF) model coupled with FLEXPART, a Lagrangian particle dispersion model, and determining the representativeness of the mountaintop measurements in these different situations.