We all know the importance of soil as a CO2 source to the atmosphere, even in high latitude/altitude ecosystems with persistent snow cover. Before the days of automated soil respiration chambers and accessible eddy covariance technology, CO2 fluxes through porous media were estimated using vertical CO2 concentration profiles coupled with assumption that the vertical profile is a direct result of the steady-state CO2 production. This approach yields a vertical diffusive flux of CO2 from the soil to the atmosphere. In the case of a snowpack this approach yields a particularly simple result, a linearly decreasing profile between the soil surface (bottom of the snowpack) and the atmosphere (top of the snowpack). Because of its relative simplicity these gradient methods are still commonly used, especially in places where eddy covariance or soil chambers are impractical or too expensive.

Unfortunately, the steady-state assumption is false more often than it is true, especially in places with high winds or uneven obstacles or terrain features. The result of the interactions between the wind and these site characteristics can cause ‘pressure-pumping’, which in turn drives CO2 concentrations in the upper meter or so of the profile away from the more-or-less linear diffusive profile to a more vertically uniform profile that roughly equals the ambient atmospheric CO2 concentration. Although wind is often associated with pressure pumping, the real driver is any phenomena that causes the ambient atmospheric pressure at the surface of a snowpack or soil to change. Further, even tiny (1 Pa or 0.01 mbar ) pressure changes can influence the near surface profile of CO2 within the soil. In general, pressure-pumping results in advective (rather than diffusive) fluxes from the porous media. The resulting changes in CO2 concentrations can be weak and intermittent or, as shown in the following visualization, they can also be extremely dynamic, persist for several minutes and hours and yield dramatically different CO2 concentrations more than a meter deep into the profile.

This pressure pumping movie, which clearly shows the effects of non-steady state pressure pumping, was compiled from snowpack CO2 profile data obtained at GLEES (US-GLE) every thirty minutes during the five days March 16-20, 2010 (for a total of 60 frames). The red asterisk in the upper right corner marks the beginning of the movie and below the date and time is an arrow, the length of which indicates the strength of the half-hourly wind speed. The vertical axis is normalized depth of the snowpack (i.e., z/h; where the snowpack depth, h, was approximately 1.8 m at the time of the movie). The horizontal axis is the CO2 mixing ratio (ppmV). The black dots are the measured profile, and the pink shaded area shows the uncertainty bounds of the model curve-fit used to describe the profile and estimate the diffusive and advective components of the total CO2 flux entering the bottom of the snowpack and exiting the top of the snowpack. Careful examination of the wind speed and shape of the profile will suggest that pressure pumping is driven by more than just wind speed – implicating uneven terrain features and/or transient atmospheric pressure variations independent of the site. The shape of the profile also indicates the partitioning between advective and diffusive fluxes. Near the top of the snowpack the profile is nearly vertical, the vertical gradient is small and advective fluxes dominate the diffusive fluxes. But near the bottom of the snowpack just the opposite is true. Here the vertical CO2 gradient is still strong, so the diffusive fluxes dominate the advective fluxes. Finally, this 5-day movie is only one example of pressure pumping that we observed during the seven years of observations. This alone is noteworthy because it strongly suggests
that pressure pumping is ubiquitous and, by inference, implies that pressure pumping can influence other soil-snowpack-atmosphere trace gas fluxes as well.

Ubiquity aside, the major result of this research is higher confidence in CO2 fluxes from the snowpack surface and the ability to detect significant year-to-year fluctuations in wintertime soil respiration. In addition, this unique study makes the GLEES wintertime CO2 flux data set one of the longest and most comprehensive of its kind anywhere. For further results of this study and the details concerning the assumptions and methods underpinning them, see Berryman et al. (2018).

References
Berryman, E. M., Frank, J. M., Massman, W. J. , and Ryan, M. G. (2018). Using a Bayesian
framework to account for advection in seven years of snowpack CO2 fluxes in a mortality-
impacted subalpine forest Agricultural and Forest Meteorology, 249, 420–433.
https://doi.org/10.1016/j.agrformet.2017.11.004.

No Comments

Be the first to start a conversation

Leave a Reply

You must be logged in to post a comment.