European greenhouse gas emissions estimated from continuous atmospheric measurements and radon 222 at Mace Head, Ireland

Sébastien Biraud, Philippe Ciais, Michel Ramonet, Peter Simmonds, Victor Kazan, P. Monfray, Simon O’Doherty, T. Gerard Spain, S. G. JenningsJournal of Geophysical Research Atmospheres 2000 · DOI

TL;DR

This paper presents a novel method for estimating European greenhouse gas emissions (CO2, CH4, N2O, CFCs) by using continuous atmospheric measurements at Mace Head, Ireland, and leveraging radon-222 (²²²Rn) as a natural tracer. The study provides independent flux estimates that largely agree with existing inventories, with a notable exception for CFC-11, suggesting potential overestimation in current inventory data.

What Problem Does This Paper Address?

Accurately quantifying regional greenhouse gas emissions is crucial for understanding their impact on climate change and for developing effective mitigation strategies. While global budgets are somewhat constrained, significant uncertainties remain regarding the magnitude and distribution of regional sources, particularly for gases like CO2, CH4, and N2O, which have both natural and anthropogenic sources. Existing methods often rely on inventories, which can have their own limitations, or on complex atmospheric transport models. This paper aims to provide an independent method for estimating European emissions by directly inferring fluxes from atmospheric measurements, overcoming some of the limitations of traditional approaches.

Methods

The core of this research lies in using radon-222 (²²²Rn) as a reference tracer. ²²²Rn is a radioactive gas with a relatively uniform continental source and a known decay rate. The authors collected continuous atmospheric data for CO2, CH4, N2O, CFC-11, CFC-12, and CFC-113 at the Mace Head station in Ireland.

The key methodological steps include:

  • Data Selection: Identifying "synoptic events" where air masses originated from western Europe, characterized by elevated concentrations of the target gases and ²²²Rn compared to marine background levels. This involved filtering based on wind direction, wind speed, and crucially, using ²²⁰Rn (a shorter-lived radon isotope) to help distinguish between continental and local (Irish) sources.
  • Correlation Analysis: Calculating the correlation between the concentration changes of each greenhouse gas (ΔCₓ) and the concentration changes of ²²²Rn (Δ²²²Rn) during these selected synoptic events.
  • Flux Estimation: Using the derived slopes from the correlation analysis and known physical properties of ²²²Rn (flux, decay rate), the authors estimated the flux densities of the target greenhouse gases. They employed two primary methods for calculating the slope:
  • Method 1: Linear regression of hourly observations of ΔCₓ versus Δ²²²Rn for all retained events.
  • Method 2: Regression of the maximum concentration values for each event.
  • Method 3: A more selective approach that retained only events with a high correlation (r > 0.75) between the species and ²²²Rn, aiming to capture more homogeneous air masses from intense emission areas.
  • Uncertainty Analysis: The study acknowledges and discusses various sources of uncertainty, including the precise source area, variations in transit time, and the non-uniformity of ²²²Rn flux across continents.

Datasets

  • Atmospheric Measurements: Continuous hourly measurements of CO2, CH4, N2O, CFC-11, CFC-12, CFC-113, ²²²Rn, and ²²⁰Rn collected at the Mace Head atmospheric research station in Ireland between January 1996 and December 1997.
  • Inventory Data: Existing emission inventories for CO2, CH4, N2O, and CFCs, including CORINAIR 1990 data and specific CFC emission inventories (McCulloch et al., 1994; McCulloch and Midgley, 1998).

Key Insights and Ideas

  • Radon as a Tracer: The study successfully demonstrates the utility of ²²²Rn as a robust tracer for inferring continental emissions, particularly when local sources can be effectively screened out.
  • Distinguishing Local vs. Continental Sources: The use of ²²⁰Rn as a tracer for local Irish sources was a critical step in isolating signals from broader western European emissions.
  • CFC-11 Discrepancy: The significant difference between the estimated CFC-11 flux and inventory data is a major finding. This suggests that current CFC-11 inventories might be overestimating emissions, possibly due to assumptions about the release rates from dumped equipment or outdated geographical distribution data.
  • Seasonal vs. Annual Fluxes: The paper highlights the importance of considering seasonal variations, especially for CO2, where winter emissions are dominated by fossil fuel use and soil respiration, while summer fluxes are influenced by biospheric uptake.
  • Methodological Robustness: The comparison of different regression methods (Method 1, 2, and 3) and the sensitivity analysis to event selection criteria provide confidence in the overall approach.

Key Results

  • Flux Estimates: The study provides estimated European flux densities for the period 1996/1997:
  • CO2 (Wintertime): 45-30 x 10³ kg C km⁻² month⁻¹
    • CH4 (Annual Mean): 4.8-3.5 x 10³ kg CH4 km⁻² yr⁻¹
    • N2O (Annual Mean): 475-330 kg N2O km⁻² yr⁻¹
    • CFC-11 (Annual Mean): 2.5-1.8 kg CFC-11 km⁻² yr⁻¹
    • CFC-12 (Annual Mean): 4.2-2.9 kg CFC-12 km⁻² yr⁻¹
  • Agreement with Inventories: The estimated fluxes for CH4, N2O, and CFC-12 were in good agreement with existing inventory data.
  • CFC-11 Underestimation: The estimated CFC-11 flux was approximately five times lower than inventory estimates.
  • CFC-12 Emission Decrease: The study suggests a decrease in CFC-12 emissions between 1996 and 1997.
  • Method 3 Insights: The more selective Method 3 yielded higher flux estimates for CFCs and CH4, suggesting it captures signals from more concentrated emission regions.

Limitations and Future Work

  • Uncertainty in Transit Time: The exact transit time of air masses from source regions to Mace Head is difficult to determine precisely, leading to a range of flux estimates.
  • Spatial Heterogeneity of ²²²Rn Flux: While assumed uniform, ²²²Rn flux can vary spatially, introducing some uncertainty.
  • CFC-113 Data: The small amplitude of CFC-113 synoptic peaks made reliable flux estimation difficult.
  • Local Source Influence: While efforts were made to minimize local influence, some residual contribution from Irish sources cannot be entirely ruled out for all species.
  • Future Work: The authors suggest that further refinement could involve using more sophisticated transport models to better constrain transit times and source regions. Investigating the spatial patterns of CFC-11 emissions and their changes over time would be crucial to reconcile the discrepancy with inventory data. Expanding this methodology to other sites could provide a more comprehensive picture of European emissions.

Summary updated

Abstract

Flux estimates of CO 2 , CH 4 , N 2 O, and CFCs over western Europe have been inferred from continuous atmospheric records of these species at the atmospheric research station of Mace Head, Ireland. We use radon ( 222 Rn) which has a fairly uniform source over continents as a reference compound to estimate unknown sources of other species. The correlation between each species and 222 Rn is calculated for a suite of synoptic events that have been selected in the Mace Head record over the period 1996/97. In the following, we describe the method and its uncertainties, and we establish data selection criteria that minimize the influence of local sources over Ireland, in the vicinity of the station, in order to select synoptic events originating from western Europe. We estimate western European flux densities of 45–30 10 3 kg C km −2 month −1 during wintertime for CO 2 , of 4.8–3.5 10 3 kg CH 4 km −2 yr −1 , 475–330 kg N 2 O km −2 yr −1 , 2.5–1.8 kg CFC‐11 km −2 yr −1 for CFC‐11, and 4.2–2.9 kg CFC‐12 km −2 yr −1 for CFC‐12. Our estimates are independent, although in good agreement with those produced by inventories, except for CFC‐11 where our estimate is much lower than the inventory.
BibTeX
@article{biraud2000european,
  author = {Sébastien Biraud and Philippe Ciais and Michel Ramonet and Peter Simmonds and Victor Kazan and P. Monfray and Simon O’Doherty and T. Gerard Spain and S. G. Jennings},
  title = {{European greenhouse gas emissions estimated from continuous atmospheric measurements and radon 222 at Mace Head, Ireland}},
  journal = {Journal of Geophysical Research Atmospheres},
  year = {2000},
  doi = {10.1029/1999jd900821},
}
Figure