Temporal variability of ammonia fluxes from open-path, eddy covariance measurements

Mark Zondlo1, Da Pan2, Katherine B. Benedict3, Bret A. Schichtel4 and Jeffrey L. Collett Jr.5

Ammonia (NH3) is an important gas phase precursor species for aerosol formation and growth, and particle phase ammonium is a large part of aerosol molar composition. Ammonia emissions to the atmosphere have high spatiotemporal variabilities. Quantifying NH3 fluxes are further complicated by severe measurement problems including efficient adsorption to instrument surfaces, low mole fractions (pptv-ppbv), and gas-particle phase partitioning. To these ends, we have developed an open-path, eddy covariance NH3 instrument that minimizes many of these sampling problems. A quantum cascade laser probes the fundamental absorption band of NH3 for high sensitivity and is tuned to an absorption feature of NH3 that retains the sensitivity while minimizing spectral interferences from other trace gas absorbers (e.g. CO2, H2O). The open-path configuration avoids the sampling artifacts associated with closed-path systems and thereby allows for fast response flux determination. Calibrations with gas standards and permeation tubes were conducted in the laboratory using direct absorption spectroscopy, and intercomparisons with commercial instruments were conducted in the field at moderate and high concentrations (> ppbv). The instrument precision is 150 pptv (10 Hz), and cross covariance analyses between vertical wind and NH3 indicate a minimal detectable flux of ~1 ng NH3/m2/s at 10 Hz. Cospectral analyses with other scalars such as temperature, moisture, methane, and carbon dioxide indicate the system has a comparable performance as other eddy covariance sensors. Measurement accuracy is ±(15% + 0.3 ppbv) with the zero point drift dominating the uncertainty in the field. Eddy covariance measurements for a three week period in summer 2015 in Rocky Mountain National Park, Colorado, showed the capabilities of the system to measure fluxes in both clean and moderately polluted conditions. Net deposition of 3.2 ng N/m2/s occurred during upslope events from the transport of air from the agricultural/urban areas on the plains. In contrast, during downslope events when clean, free tropospheric air flowed across at the site, re-emission of NH3 to the atmosphere of a similar magnitude was observed. The system is being re-deployed at RMNP for the summer 2016 at the same site. Ongoing efforts to improve the zero point drift at sub-ppbv levels and variations in NH3 fluxes through the summer will be discussed.


1Princeton University, mzondlo@princeton.edu
2Princeton University, dp7@princeton.edu
3Colorado State University, kbenedic@engr.colostate.edu
4Natl. Park Service – Air Resource Division, Bret.Schichtel@colostate.edu
5Colorado State University, collett@atmos.colostate.edu