Sources and sinks of formic, acetic, and pyruvic acids over central Amazonia: 2. Wet season

Abstract

We have determined the gas phase concentrations of formic (HCOOH), acetic (CH3COOH), and pyruvic (CH3C(O)COOH) acids in the forest canopy, boundary layer, and free troposphere over the central Amazon Basin during the April–May segment of the 1987 wet season. At 150‐m altitude in the boundary layer the daytime average concentrations were 430±225, 340±155, and 25± 5 ppt for HCOOH, CH3COOH, and CH3C(O)COOH, respectively. These values were fivefold lower than those observed in the 1985 dry season. Concentrations measured near canopy top were not significantly different from boundary layer values (P = 0.10), while concentrations in the lower canopy were significantly less. Concentrations in the free troposphere (5 km) were lower than in the boundary layer and averaged 170±40, 210±40, and 15±15 ppt for HCOOH, CH3COOH, and CH3C(O)COOH, respectively. Fivefold enhancements of CH3C(O)COOH concentrations were observed in convective outflows at 5‐ to 6‐km altitudes. Aerosol carboxylate concentrations were usually below our detection limit of 5–10 ppt. Preliminary branch enclosure measurements indicated significant direct emission of carboxylic acids by vegetation. A one‐dimensional photochemical model for the canopy and the boundary layer was used to examine the contributions from various sources to the carboxylic acid budgets. Model results indicate that direct emissions from vegetation can account for most of the concentrations observed in the canopy. These emissions peak during the daytime hours, and 24‐hour average upward fluxes at canopy top are 4.4×109, 3.7×109, and 2.8×108 molecules cm−2 s−1 for HCOOH, CH3COOH, and CH3C(O)COOH, respectively. However, direct emissions from vegetation can account for only a small fraction of the observed carboxylic acid concentrations in the boundary layer, suggesting a large contribution from atmospheric sources. The atmospheric reactions previously suggested in the literature as sources of carboxylic acids (gas phase decomposition of isoprene, CH3CO3 + peroxy, aqueous phase oxidation of CH2O) appear to be too slow to explain the observed concentrations. Other atmospheric reactions, so far unidentified, could make a major contribution to the carboxylic acid budgets.