Assessing Day-to Day Variability in the Vertical Distribution of Methane, Carbon Dioxide, and Ozone over Railroad Valley, NV

In-situ observations of three trace gases over a remote desert site allow for an analysis of the variability of ozone (O3), methane (CH4), and carbon dioxide (CO2) in the free troposphere. Observations from June 2013 show almost no change from one day to the next in the boundary layer (BL) up to > 4 km (30% of the atmospheric column), while mixing ratios of methane and carbon dioxide show strong variability above this altitude. Ozone values also demonstrate variability above the boundary layer, and ozone day-to-day variability in the well-mixed BL is greater than that of CO2 or CH4. Results from week-long intensives in both June 2012 and June 2013, as well as monthly measurements over the period 2012-2013, will be compared to long-term vertical profile data sets at other locations (Trinidad Head, CA; Briggsdale, CO; and the Southern Great Plains site, OK). Variability above and in the boundary layer will be reported. To assess possible sources of variability, in situ data will be analyzed with a chemical trajectory model (GEOS-Chem v9-01-03). The North America nested-grid version of GEOS-Chem utilizes varying emission inventories and model parameterizations to simulate the emissions of greenhouse gases (CO2 and CH4, in this case) and ozone precursor gases. Tagged tracer simulations in GEOS-Chem allow for the geographical source apportionment of ozone, indicating whether the observed O3 was formed in the upper troposphere, middle troposphere, stratosphere, or any user-defined boundary layer location. For this study we will focus on ozone formed in the boundary layer over Asia, the Pacific Ocean, Mexico, Canada, and the United States. The importance of daily variability in the free tropospheric values of CO2, CH4, and O3 will be discussed in the context of column measurements collected from the surface or from space. Many data assimilation systems are designed to assume that changes to the total column average should be attributed primarily to changes within the boundary layer where large diurnal and seasonal cycles are presumed to dominate, but our data show that under some circumstances, this is not a valid presumption and can lead to a misinterpretation of the column measurement.