Investigating methods used to quantify gaseous emissions from vegetation fires using spectroscopic measurements

摘要

This work investigates the application of ground-based trace gas spectroscopy to deter- mine the chemical makeup and quantity of smoke emitted from vegetation fires. Ultraviolet Differential Optical Absorption Spectroscopy (UV-DOAS) has been infrequently deployed in fire emission studies, yet is potentially a portable, lightweight, inexpensive and simple method. Fourier Transform Infrared (FTIR) spectroscopy has been more commonly used in fire emissions studies, but not generally in the long ( 10 m) open-path ground-based geometry explored here. This research combines these approaches to investigate their ability to quantify trace gas fluxes emitted from open vegetation fires, in part to help validate estimates of fuel consumption rate based on fire radiative power [FRP] measures. UV and IR measurements of the smoke plumes from controlled open vegetation fires ( 4 hectares) were recorded during three field campaigns in Arnhem Land (Northern Australia), Kruger Park (South Africa) and Alberta (Canada). The UV-DOAS was used to quantify NO2 and SO2 vertical column amounts (maximum column amounts approx 200 ppmm), allowing the determination of flux-rates when used to traverse the smoke plume and coupled with plume velocity estimates. Horizontal column amounts of the main plume carbonaceous species (CO2, CO and CH4) were quantified using FTIR methods and used to calculate emission ratios and emissions factors for the target gases, providing detail on inter- and intra- fire variations that are often available from the current literature. Providing NO2 and SO2 are detectable by the FTIR, UV-DOAS flux-rates and FTIR emissions ratios can be combined to calculate flux rates for all FTIR-detectable species. This allows for the determination of the total carbon flux from the fires, and its variation over time. Since vegetation is approximately 50% carbon, this flux is in theory directly proportion to the fuel consumption rate, and directly comparable to the fire’s radiative power output variations as determined by airborne thermal imaging. Hence, in addition to providing the means to estimate smoke plume chemical makeup, emissions magnitude and variability, the simultaneous deployment of the techniques of UV-DOAS, FTIR spectroscopy and airborne thermal imaging enables the validation of FRP derived fuel consumption rates. The FRP method is gaining ground as a tool for improving biomass burning emissions inventories based on satellite observations, but at present has had relatively little validation. This study therefore contributes to the ongoing evaluation effort. Findings demonstrate that the UV-DOAS is an effective way to measure column amounts of SO2 and NO2 in vegetation fire plumes, providing that the fires are of an adequate size and emit smoke in sufficient quantities. The exact nature of the ability to accurately quantify NO2 and SO2 using the method did have a dependence on fuel type, since the combustion of different fuel types (e.g. grasses vs. woody fuels vs. organic soils) appeared to cause more of less of these particular gases to be emitted. There was difficulty in confidently detecting NO2 via the OP-FTIR approach for the majority of the study cases, due to the relatively weak IR absorption bands used and the relative scarcity of this gas in the plumes in comparison to some others studied. We advocate using the UV-DOAS and FTIR combination in relation to trace gas measurements from vegetation fires, providing SO2 or NO2 can be identified by the FTIR in the particular biomass burning situation under study. Where simultaneous FRP measurements are available, the carbonaceous flux rates calculated using the FTIR/UV-DOAS method show a strong correlation with FRP, helping to confirm the relationship between FRP and fuel consumption rate at the scale of these vegetation fires. This is to our knowledge currently by far the largest fires upon which this relationship has been evaluated, prior evaluations being limited to laboratory-scale events only.