Atmospheric Environment: X (Oct 2021)

Characteristics of organic components in PM2.5 emitted from peatland fires on Sumatra in 2015: Significance of humic-like substances

  • Yusuke Fujii,
  • Susumu Tohno,
  • Hiroki Kurita,
  • Haryono Setiyo Huboyo,
  • Badrus Zaman

Journal volume & issue
Vol. 11
p. 100116

Abstract

Read online

We characterize fine particulate matter (PM2.5) emitted from Indonesian peatland fires using ground-based source-dominated samplings of PM2.5 near peatland fire sources at two Regencies in Riau, Sumatra, Indonesia (number of samples = 13). Organic carbon (OC), elemental carbon, water-soluble OC (WSOC), the carbon content of humic-like substances (HULIS-C), and biomass burning tracers are determined. The carbon mass ratios of WSOC to OC (0.085 ± 0.015) and HULIS-C to WSOC (0.55 ± 0.085) are fairly constant and independent of the peatland fire sampling sites. By comparing diagnostic ratios using OC, WSOC, and HULIS-C at the peatland fire source and the receptor site (Malaysia) during peatland fire-induced haze periods, secondary WSOC and HULIS-C formation during transport from the source to the receptor site is highly possible. Interestingly, the mass ratio of syringic acid to levoglucosan (0.045 ± 0.0075) is fairly constant at Indonesian peatland fire sources. Because syringic acid is less stable than levoglucosan, this ratio is an aging indicator for Indonesian peatland fires at receptor sites. By comparing the mass fraction of each organic compound in the present study and previous studies, it is evident that the source profile for the coburning of peat with surface vegetation is significantly different compared with the burning of peat alone. Further knowledge of peat burning emissions is needed, particularly with respect to burning conditions, peat composition, and the effects of vegetative burning on peatland. Improved knowledge of these factors would lead to more reliable speciated emission inventories of Indonesian peatland fires, advancing chemical transport and radiative forcing modeling, as well as health risk assessment.

Keywords