Table of Contents:
Atmospheric Measurement of Greenhouse Gas Fluxes
Unlike observation strategies useful for direct GHG flowrate measurement, like CEMs, flux quantification for GHG fluxes originating from surface processes where direct measurements are not possible must be accomplished using atmospheric observations of mole fraction coupled with inferential analysis methods. As would be expected, these require detailed information describing atmospheric properties and motions in addition to greenhouse gas concentration observations. Atmospheric motions at fine spatiotemporal scales are realized with simulation models, whether of the global atmosphere or more locally focused numerical weather prediction (NWP) models. Atmospheric parameters, such as wind speed and direction, temperature and pressure, are needed tlnougliout the estimation and modelling domain. These, coupled with mole fraction observations at various domain locations, provide a means to track atmospheric greenhouse and other observable trace gases through the atmosphere. From the carbon management point of view, these methods allow for the estimation of source/siuk locations and the magnitude of the flux at those locations.
Equation (5) relates greenhouse gas, or any relatively long-lived trace gas, atmospheric mass flux to observable parameters, or, in the case of air density, to a parameter directly computed from observed parameter values. Mass flux is a r ector quantity due to its dependence on velocity. Both must be represented by a magnitude and direction.
where: MG = Mass flux of the greenhouse gas of interest,
xghg = Greenhouse gas mole fraction,
MA = Mass flow or flux of the atmosphere,
p t = Density of the atmosphere, and
VA = Atmospheric velocity.
Atmospheric velocity r ectors, composed of wind speed and direction values, across the geographic domain of interest impart these properties to mass flux with scalar quantities of greenhouse gas concentration and air density. Descriptions of atmospheric motions both at global scales, known as global circulation models, and at more local scales, NWP models, are used to simulate greenhouse gas flows in the atmosphere. Although these modelling approaches rely heavily on fundamental descriptions/models of the fluid dynamic system of the atmosphere, they are aided by the use of meteorological observations. In most cases, meteorological parameter measurement data is available at a relatively few points in the domain relative to the number of cells in a computational domain. Advances in and assimilation of radar- based observation, both from the surface and space, have substantially improved the skill of atmospheric property simulation from the surface to the top of the atmosphere. As discussed in more detail below, NWP and global atmospheric circulation models can provide simulated atmospheric parameter values throughout a domain of interest with varying degrees of temporal and spatial resolution. These, coupled with GHG mole fraction measurement data, are used in the inferential method described below for tracking atmospheric GHG flows from their sources on or near the surface, through the planetary boundary layer, and to the global atmosphere.
Greenhouse gas dynamics in the global atmosphere
The movement of CO, and methane at continental and global scales has been estimated with global atmospheric circulation models, coupled with mole fraction measurements, through the contributions of several nationally-sponsored, mole fraction observing systems (ECMWF, 2018; Rodeuberg. 2017). The CarbonTracker (CT) model, developed by NOAA's GMD (NOAA-CT, 2017), estimates atmospheric carbon uptake and release globally. The observed mole fractions used are predominantly taken from northern hemisphere locations, which tends to make simulated values there somewhat more accurate. CT, and the associated long-term monitoring of atmospheric CO, upon which it depends, are tools to advance the understanding of carbon uptake and release from land ecosystems and the oceans. Providing a means of studying the movement of atmospheric CO, globally, CT and other global models can be tools for monitoring environmental changes, including human management of land and oceans and the potential impact of carbon management efforts globally. A companion model tracking global methane emissions (NOAA-GMD, 2018). provides methane emission flux simulations from North American and global natural and anthropogenic methane emissions. Other global-scale GHG inversion models also produce flux estimates annually (Wageningen Univ. & ICOS Netherlands, 2019; ECMWF-CAMS. 2019; Max Planck Institut for Biogeochemistiy, 2019). Many assimilate satellite observations either in addition to in situ data or alone. Various Bayesian data assimilation methods are utilized, including Ensemble Kalman Filter and 4D-Var methods.
The CarbonTracker system is an example of atmospheric inversion analysis combining statistical optimization, global atmospheric circulation simulation, and CO, and CH, observational data. The underlying global transport models have similarity to finite element, finite difference analyses used widely in many science and engineering applications. These methods enable the investigation of the behavior of complex systems using spatial and temporal gridding of the domain containing the item of interest and embedding fundamental process descriptions or parameterizations of them within each grid cell. For example, a spatial grid encompassing a body where thermal and or mechanical stresses are to be investigated or that of a fluid in a flow field confined to a specified geometry, pressure and temperature regime can be simulated and compared with measured parameters. In meteorological models, three- dimensional grids beginning at Earth's surface and extending to the top of the atmosphere are used to simulate atmospheric dynamics and properties over specific geographical regions and time periods. Physical and hydrodynamic principles and parameterizations of individual Earth system processes, e.g., cloud formation, planetary boundary layer dynamics, land-atmosphere gas exchange, and solar radiation intensity at the surface, are used at each grid cell along with pertinent input data. At each tune step, the model provides an array of atmospheric parameter values, e.g., pressures, temperatures, wind direction and speed, and surface energy exchange. As described below, application of NWP models to simulate atmospheric dynamics at local scales is a tool to investigate greenhouse motions and source and sink characteristics in urban areas. These are important to carbon management as the smaller geographic regions of cities (small relative to the total laud surfaces of the earth) har e outsized contributions to national and global emissions, and, therefore, are likely to be a focus of carbon management efforts.