Using stable carbon and oxygen isotopes to attribute measured carbon-dioxide emissions in urban environments to different fuel sources.

Transcription

1 Using stable carbon and oxygen isotopes to attribute measured carbon-dioxide emissions in urban environments to different fuel sources. Andreas Christen (1), Joseph K. Lee (1), Rick Ketler (1), Zoran Nesic (1,2), Luitgard Schwendenmann (3), Caitlin Semmens (1) (1) Department of Geography, The University of British Columbia, Vancouver, BC, Canada (2) Biometeorology and Soil Physics Group, The University of British Columbia, Vancouver, BC, Canada (3) School of Environment, The University of Auckland, Auckland, New Zealand Vancouver (Photo: A. Christen)

2 Motivation The recent decade has seen a rapid adoption and advancement of methods to interpret direct measurements of carbon dioxide (CO 2 ) in urban environments to provide independent carbon emission estimates at urban scales. Although several studies demonstrate potential of using direct measurements to validate fine-scale emission inventories of CO 2, a major challenge remains the source attribution of total measured mass fluxes or concentrations of CO 2 to various sources.

3 The challenge of source attribution Measured net mass flux of carbon dioxide FCO2 = C + R - P + ΔS A partitioning of FCO2 into the different fuel and biogenic sources is not possible using current approaches.

4 Research question Can stable isotope composition of CO 2 add additional information on fuel sources and hence complement emission and concentration measurements of CO 2 in urban environments? The 3 most abundant isotopologues of CO 2 in the atmosphere are: 13 C 16 O 2 12 C 16 O 2 12 C 16 O 18 O ~98.43% ~1.10% ~0.39% ~4 ppm ~400 ppm ~1.5 ppm 13 C 18 O

5 Stable isotope ratios We express the relative abundance (ratios) of the different isotopologues in delta-notation: 13 [ 13 C 16 O 2 ]/[ 12 C 16 O 2 ] C = R VPDB Ratio of a pre-defined standard sample 18 [ 12 C 16 O 18 O]/[ 12 C 16 O 2 ] O = R SMOW Ratio of a pre-defined standard sample

6 Detectors Sample cell with intake from atmosphere Reference cell with known gas Beam splitter We can measure CO 2 - isotopologues directly in-situ in the urban atmosphere using a spectroscopy system (here: TGA 200, Campbell Scientific Inc., Logan UT). Tunable laser

7 Laser spectroscopy of CO 2 isotopologues Wavenumber (cm -1 ) Detector Signal (mv) CO cm CO cm CO cm CO 18 O cm Laser Current (ma) 12 CO cm At low pressure, absorption bands separate

8 Objectives 1. Determine the typical δ 13 C and δ 18 O of various emission sources contributing to CO 2 in the urban boundary layer over Vancouver, Canada. 2. Determine the CO 2 enhancement in Vancouver s urban boundary layer. 3. Measure the δ 13 C and δ 18 O of the enhancement directly and determine whether the added (i.e. enhanced) CO 2 reflects the isotopic signatures of the typical CO 2 sources expected.

9 Major fuel sources separate well! -5 Exhaust samples -10 Gasoline δ 18 O ( VPDB-CO 2 ) Natural gas Diesel no catalytic converter TEDLAR bags δ 13 C ( VPDB-CO 2 ) Dilution and analysis in lab

10 However soil and plant respiration overlap with gasoline 0 5 Plant and soil respiration Atmospheric background δ 18 O ( ) 15 Gasoline Human respiration Diesel 20 Natural Gas δ 13 C ( )

11 Year-round measurements on UBC campus Downtown Ocean 3m, urban background, 200m from nearest road Metro Vancouver (Photo: A. Christen)

12 Urban CO 2 enhancement 460 Winter (Dec Feb) 460 Spring (Mar May) CO2 Mixing Ratio (ppm) Vancouver (Urban) ppm 20.4 ppm ppm Estevan Point (Pristine) ppm 11.6 ppm ppm Summer (Jun Aug) 460 Fall (Sep Nov) CO2 Mixing Ratio (ppm) ppm 15.0 ppm ppm ppm 14.4 ppm ppm Time of day (LST) Time of day (LST) Annual average difference: ppm ppm = 16.4 ppm

13 CO 2 mixing ratio (ppm) Night Day Night Mixing ratio Vancouver UBC to Tropospheric Background ppm Tropospheric background Isotope ratio for 13 C δ 13 C ( ) Hour of day (PST) 11. 0

14 Keeling plot for a 24 hour period ppm Vancouver UBC 9 δ 13 C (per mil) Winter January 19, /[C O 2 ] (ppm 1 )

15 Keeling plots for a day in winter and spring ppm Vancouver UBC 9 Spring May 5, 2013 δ 13 C (per mil) Winter January 19, /[C O 2 ] (ppm 1 )

16 Community Energy and Emissions Inventory 41.7% Natural gas 5.26 Mt CO2e yr % Gasoline Vehicles Buildings kt CO2e month -1 1, % Diesel Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

17 Comparing CEEI and Isotopes January May Observation based on δ 13 C of CO2 in urban atmosphere 53% Natural gas 47% Gasoline Diesel Respiration 25% 75% Inventory Monthly distributed CEEI / BEM for entire metropolitan area 58% Natural gas 42% Gasoline Diesel 31% 69%

18 0 5 wind from city wind from ocean δ 18 O δ 13 C 0º 90º 180º 270º 360º wind direction (º from North)

19 Intercepts change over year Wind from city sector only 0 δ 18 O (per mil) δ 13 C Week of Year In winter, when there is home heating (natural gas), there is lower d 13 C and d 18 O from city sector

20 Intercepts are temperature dependent 5 Weekly Colder temperatures mean more heating using natural gas (lower δ 13 C) 15 δ 18 O (VPDB-CO 2 ) δ 13 C Weekly average air temperature(ºc)

21 Intercepts change over day and year 10 Dec/Jan 10 Feb/Mar 10 Apr/May δ 18 O = 11.6 (VPDB-CO 2 ) (VPDB-CO 2 ) δ 18 O = 16.9 δ 18 O = δ 13 C = 30.9 δ 13 C = 32.2 δ 13 C = Jun/Jul 0 δ 18 O = δ 13 C = Aug/Sep 0 δ 18 O = δ 13 C = Oct/Nov 0 δ 18 O = δ 13 C = Local standard time δ 13 C Gasoline Local standard time δ 13 C Natural gas Local standard time

22 Annual changes in isotopic signatures δ 18 O ( VPDB-CO 2 ) Unusually cold Diesel Gasoline 25 Natural gas δ 13 C ( VPDB-CO 2 )

23 Summary and conclusions The urban boundary layer (UBL) of Vancouver is enriched by CO 2 by on average +15 ppm. Natural gas separates well from all other sources with its low δ 13 C (-41.6 ) and low δ 18 O (-22.7 ). Atmospheric measurements of δ 13 C confirm that the UBL contains a higher fraction of CO 2 from natural gas in winter and night, and more gasoline / diesel during summer and day, consistent with inventories. Challenges: The δ 13 C of gasoline (-27.3 ) is close to diesel (-28.8 ) and overlaps with respiration. δ 18 O has large variations. Continuous measurements of isotopic composition have potential to complement emission estimates at urban scales.

24 Challenges and next steps Understand more carefully what controls δ 18 O variations in gasoline exhaust and respiration and how they vary over the year. Explore a 3-end-member mixing model using δ 13 C and δ 18 O on seasonal and diurnal scales. More carefully characterize regional background. Future: Develop eddy covariance measurements of CO 2 isotopologues to characterize fleet signatures etc. We acknowledge funding and support from