Seasonal, Day-of-Week, and Diurnal Trends in CO2 and CH4 Observed at the Sandia Livermore Tower in Livermore, CA USA
Abstract:Urban centers contribute a majority of global anthropogenic greenhouse-gas (GHG) emissions and are therefore a natural target for policies aimed at GHG emissions reductions for climate change mitigation. Emissions inventories are often used to assess the impact of policies implemented to reduce GHG emissions, but these bottom-up approaches are plagued by large uncertainties. Top-down approaches are therefore necessary to verify the results and quantify and reduce the uncertainties.
Here we present a full year of continuous in situ measurements of CO2 and CH4 from a site in Livermore, CA, which is well positioned to capture the outflow from a large metropolitan region that includes San Francisco, the East Bay, and San Jose. These measurements are ongoing and are made using a Picarro CO2/CH4/H2O analyzer sampling at 27 meters a. g. l. at the Sandia Livermore tower (Lon: -121.71˚, Lat: 37.67˚). The measurements are calibrated ~daily using whole air cylinders that are referenced to the NOAA/WMO scales. From these calibrations, the long term stability of the instrument was determined to be 0.032 ppm for CO2 and 0.24 ppb for CH4. As part of our discussion, we will describe the sampling platform, the automated calibration system, the various post-processing QA/QC procedures used, and some analytical figures of merit.
We have analyzed the ambient mole fraction data, referenced to a clean air background, across seasonal, day-of-week, and diurnal time scales and as a function of various meteorological parameters, including wind direction and boundary layer height in order to isolate different sources of variability in the CO2 and CH4 signals. We have found, for example, that substantial weekday/weekend differences exist in the CO2 signal (higher on weekdays) but not in the CH4 signal, consistent with the expectation that the transportation sector is a stronger contributor to CO2 concentrations than to CH4 concentrations in the region. We have also compared the observations to hourly simulations of CO2 and CH4 mole fractions from a Lagrangian Particle Dispersion Model (FLEXPART-WRF) combined with emissions inventories of both gases, analyzing the simulated data for similar temporal patterns to those found in the observed data.