Atmospheric methane isotopes identify inventory knowledge gaps in the Surat Basin, Australia, coal seam gas and agricultural regions
- 1School of Biological, Earth and Environmental Sciences, The University of New South Wales (UNSW Sydney), NSW, 2052, Australia
- 2MetAir AG, Airfield LSZN, Switzerland
- 3Airborne Research Australia, Parafield Airport, SA, 5106, Australia
- 4College of Science and Engineering, Flinders University, SA, 5001, Australia
- 5Environmental Defense Fund, Third Floor, 41 Eastcheap, London, EC3M 1DT, United Kingdom
- 6Department of Earth Sciences, Royal Holloway, University of London, Egham, TW20 0EX, UK
- 7Institute for Marine and Atmospheric research Utrecht (IMAU), Utrecht University, Utrecht, 3584 CC, the Netherlands
Abstract. In-flight measurements of atmospheric methane (CH4(a)) and mass balance flux quantification studies can assist with verification and improvement of UNFCCC National Inventory reported CH4 emissions. However, attributing CH4(a) mole fraction readings to one or more emission sources is difficult where co-located plumes mix rapidly within the convective boundary layer. The stable carbon isotopic signatures of CH4 sources can assist with source attribution and potentially identify bottom-up (BU) inventory knowledge gaps and identify mitigation opportunities. In the Surat Basin, Queensland, Australia, both coal seam gas (CSG) production and cattle farming are increasing sources of CH4 emissions into the atmosphere. The CSG fields cover thousands of square kilometres and many CSG facilities and farms are inaccessible as part of ground-based surveys.
The aims of this study were to explore the use of CH4(a) isotopic composition (δ13CCH4(a)) of in-flight atmospheric air (IFAA) samples and the application of source endmember mixing models to assess where the BU inventory was well characterised, to identify potential gaps in the BU regional inventory (missing sources, or over- and underestimation source categories), identify mitigation opportunities, and to characterise the isotopic signature of CH4 sources that were inaccessible during ground surveys. Forty-nine useable IFAA samples were collected between 100–350 m above ground level (mAGL) over a region of CSG, coal mining and agricultural production over a 2-week period in September 2018. For each IFAA sample the 2-hour back trajectory footprint area was determined using the NOAA HYSPLIT atmospheric trajectory modelling application. Samples were then assigned to a source category set where over 50 % of the 2-hour upwind BU inventory could be attributed to a single source (CSG, grazing cattle, or feedlots), and further differentiated based on altitude (100–200 mAGL or 250–350 mAGL). A novel multi-dataset (source category) regression method with shared parameters (background air CH4(b) and δ13CCH4(b)) was used to fit Keeling models simultaneously to each data set. The determination of a common background endmember (CH4(b) and δ13CCH4(b)) for the two-endmember mixing model reduces the uncertainty in the derived isotopic signature for a source (δ13CCH4(s)). The estimated δ13CCH4(s) signatures for each category were then compared to the database of source δ13CCH4(s) signatures established from ground studies.
For IFAA samples collected between 250–350 mAGL altitude, the best-fit δ13CCH4(s) signatures compare well with the ground observation: CSG δ13CCH4(s) −55.5 ‰ (CI 95 % ± 13.4 ‰) versus δ13CCH4(s) −56.7 ‰ to −45.6 ‰; grazing cattle δ13CCH4(s) −60.5 ‰ (CI 95 % ± 15.1 ‰) versus −61.7 ‰ to −57.5 ‰. For feedlots, the derived δ13CCH4(s), −69.5 ‰ (CI 95 % ± 22.1 ‰), was isotopically lighter than the ground-based study (δ13CCH4(s) from −65.2 ‰ to −60.3 ‰), but within agreement given the large uncertainty for this source. For IFAA samples collected between 100–200 mAGL the δ13CCH4(s) signature for the CSG set, −65.3 ‰ (CI 95 % ±13.1 ‰), was isotopically lighter than expected, suggesting a BU inventory knowledge gap or the need to extend the population statistics for CSG δ13CCH4(s) signatures. For the 100–200 mAGL set collected over grazing cattle districts the δ13CCH4(s) signature, −52.5 ‰ (CI 95 % ± 18.8 ‰), was much heavier than expected from the BU inventory. This is likely due to CSG CH4 emissions entering the study domain from an adjacent CSG field. Using the multi-Keeling-model regression derived background air values (1.825 ppm (CI 95 % ± 0.037 ppm) and −47.3 ‰ (CI 95 % ± 0.3 ‰)), a Keeling model fitted to the isotopically light set yielded a low δ13CCH4(s) signature of −80.5 ‰ (CI 95 % ± 9.2 ‰). A CH4 source with this low δ13CCH4(s) signature and a high rate of emissions has not been incorporated into existing BU inventories for the region. Further ground-based studies are required to identify the isotopically light CH4 source, with possible sources including termites and CSG brine ponds. If the excess emissions are from the brine ponds, they can be abated. It is concluded that in-flight atmospheric δ13CCH4(a) measurements used in conjunction with endmember mixing modelling of CH4 sources are powerful tools for BU inventory verification and for identifying sources that can be mitigated. The subregions identified for BU inventory review would likely have been overlooked based on CH4(a) measurements alone.
Bryce F. J. Kelly et al.
Bryce F. J. Kelly et al.
Bryce F. J. Kelly et al.
Viewed (geographical distribution)