Submitted to: National Atmospheric Deposition Program's Ammonia Workshop
Publication Type: Abstract Only
Publication Acceptance Date: 9/29/2003
Publication Date: 10/23/2003
Citation: Luke, W.T., Meisinger, J.J., Schomburg, C.J., Scudlark, J.E., Siefert, R.L., Larson, R.K., Heard, I.K. 2003. A field intercomparison of eight ammonia measurement techniques [abstract]. National Atmospheric Deposition Conference. p. 1. Interpretive Summary:
Technical Abstract: From November 28 to December 10, 2001 the responses from eight active and passive sampling techniques designed to measure ambient gas phase ammonia (NH3) and particulate ammonium (NH4+) were compared at a field site near Washington, DC which represented a predominantly agricultural ammonia environment. The intercomparison site was 150 m from a free-stall dairy barn facility at USDA's Beltsville Agricultural Research Center (BARC) in Beltsville, MD. The inlets of all the samplers were located within 3 m horizontally and 1 m vertically from one another, at a height of approximately 4 m. The comparison consisted of a total of eighteen sampling intervals, ranging from 6 to 23 hours. The techniques compared included four active, batch sampling methods (cyclone separator and annular denuder/backup filter collection of NH3 and NH4+ followed by ion chromatography (IC); honeycomb denuder/filter collection of NH3 and NH4+ followed by IC; phosphoric acid solution impinger/IC measurements of total ammonium; and bulk filterpack collection of NH3 and NH4+ followed by IC analysis). Two continuous techniques were also compared: a mist chamber to collect total ammonium and p-NH4+ followed by fluorescence derivitization with a liquid wavecore detector; and a commercial chemiluminescence detector. Finally, two passive sampling devices for gas-phase NH3 were deployed: a Willems badge design and an Ogawa passive sampler. A more limited comparison of the annular denuder system with the commercial chemiluminescence detector was conducted from December 18-22, 2001. Run-averaged NH3 concentrations ranged from 0.1 to over 30 µg N/m3 during the test period. The outputs of the continuous techniques reveal considerable temporal variation in NH3, with concentrations ranging from almost zero (during fog and precipitation events) to over 150 µg N/m3 in transient events. The comparison demonstrated the inherent difficulty in accurately quantifying ambient NH3 near an active NH3 source. Agreement between some of the techniques was quite good. For example, annular denuder and bulk filterpack results agreed well (regression coefficient, m, of 0.90, r2 = 0.968, n=12) at integrated loadings of less than 10 µg N/m3 on average; at higher concentrations, however, the filterpack technique suffered from breakthrough, and consistently underestimated NH3 concentrations. Average NH3 concentrations reported by the annular denuders agreed well with those measured via chemiluminescence (m = 0.830, r2 = 0.964, n = 19). The passive Ogawa and Willems samplers reported concentrations which agreed quite well with annular denuder results (m = 1.17, r2 = 0.968, n = 3). For total NH4+, agreement between the annular denuder and the acid scrubbers showed similar response (m = 0.509), but with considerable scatter (r2 = 0.241, n = 18), possibly due to inclusion of large particulate matter in the scrubbers but not in the denuders. Results from all the techniques will be compared using paired regression analyses. In addition, the temporal behavior of [NH3] will be investigated at this location in close proximity to a major emission source.