Two ice cores, identified as D4 and D5, were collected in 2003 from high

Transcription

1 Supporting Online Material Methods Ice core collection and analysis Two ice cores, identified as D4 and D5, were collected in 2003 from high snowfall regions of the west-central Greenland ice sheet (Fig. S1). Located at high elevation (2713 m above sea level) where summer surface melting is rare, the more northern D4 core (71.4 o N, 44.0 o W) was drilled to a depth of 144 m [Banta and McConnell, 2007]. Located ~350 km to the south, the 150 m D5 core (68.5 o N, 42.9 o W) was drilled at a lower elevation (2473 m above sea level) where summer surface melting is common. Both cores were drilled with a 10 cm diameter electromechanical drill. The ~1 m long core sections were measured and weighed in the field and returned frozen to the National Ice Core Laboratory for sampling. Five parallel longitudinal samples were cut from the cores, each with a cross-section of ~3.5 by ~3.5 cm, and the samples were shipped frozen to our ultra-trace chemistry laboratory in Reno for analysis. The D4 and D5 ice cores were analyzed for a broad range of elements and chemical species using a unique continuous flow analysis system [McConnell et al., 2002a; 2002b; 2007]. This system includes a traditional continuous flow analysis stage, two high resolution inductively coupled plasma mass spectrometers, and an inductively coupled plasma optical emission spectrometer. Analytes in this study included indicators of continental dust (aluminum, magnesium, calcium, iron), sea salts (sodium, magnesium, calcium, sulfur), industrial pollution (lead, sulfur, nitrate), explosive volcanism (sulfur), and atmospheric photochemistry (hydrogen peroxide). Nearly all analytes showed a strong seasonal cycle that can be used for annual-layer

2 counting. Although seasonal cycles for a number of analytes were used to confirm annual-layer identification, here we used minima in hydrogen peroxide concentration to identify the mid-winter horizon (nominally January 1 of each year) [McConnell et al., 1997]. While hydrogen peroxide is frequently used for ice core dating in shallow cores, diffusion smoothes the annual cycle with time. At the D4 core site, the high snowfall rates and cold temperatures mean that very distinct annual cycles in hydrogen peroxide are preserved throughout the cores, providing a means to obtain consistent sub-seasonal dating throughout the record. While more diffused and disrupted by melt layers, accurate annual dating at the D5 site was accomplished using annual cycles in hydrogen peroxide, sulfur, and dust tracers such as aluminum and calcium. For first analysis of the D4 ice core, we added a triple quadrapole mass spectrometer to the continuous flow analysis system, making possible for the first time high depth resolution, very low concentration measurements of polar organic and other molecules that are exactly co-registered with all other elemental and chemical measurements. As an indicator of fallout from biomass burning emissions, we measured vanillic acid (VA). Unlike ice core tracers of biomass burning such as ammonium ion (NH4+) or nitrate (NO3-) which have significant sources other than wildfires [Holdsworth et al., 1996; Dibb et al., 1996], VA is predominantly emitted during combustion of conifer trees and so provides a tracer of conifer-rich forest fire emissions [Simoneit, 2002; Fine et al., 2004]. For a second analysis of the D4 ice core, we added a black carbon analyzer (SP2, Droplet Measurement Technologies, Boulder CO) to the continuous flow analysis system to yield the first high depth resolution, very low concentration and size distribution measurements of BC in ice cores during recent centuries. Individual BC

3 particles are counted in the method, and the counts accumulated during 30-second measurement intervals to derive BC particle number and mass distribution. The D4 ice core record contains an average of 13, 30 second measurement intervals per year from 1788 through 2002, with BC measurements exactly co-registered with all other elemental and chemical measurements. BC detection limits are ~0.02 ng g -1. The SP2 uses a patented technique [Stephens et al., 2003] that measures mass of individual BC particles using laser-induced incandescence (see [Schwarz et al., 2006] for additional details). BC particles pass through the laser beam intra-cavity of a diode-pumped Nd YAG laser (1.06 µm wavelength) where they absorb light. Circulating power in the cavity is in excess of 1 MW cm -2. BC particles absorb sufficient energy as they pass through the laser beam and reach a temperature at which they incandesce. Intensity of the incandescence is measured with a photomultiplier tube and recorded. Mass of individual BC particles is proportional to the area of the incandescence signal. The SP2 was calibrated using glassy carbon particles of known density with particle size selected using an electrostatic classifier. In this study, calibration was conducted over a particle mass range from 0.23 to 41 femptograms for individual particles.

4 # D4 # D Kilometers Fig. S1. Map of Greenland region showing locations of the D4 and D5 ice cores. Also shown is normalized residence time for the D4 site based on three-day isentropic air mass back-trajectory modeling for the period from 1958 through 2002 using only trajectories corresponding to periods of significant snow accumulation at the core site as determined from ERA-40 reanalysis. NEEDS COLOR LEGEND IN FIGURE!

5 Fig. S2. Monthly averaged BC and hydrogen peroxide (H 2 O 2 ) concentrations in the D4 ice core record from 1860 to The well preserved, strongly seasonal H 2 O 2 record, together with a range of other elements and chemical species, was used to derive the depth-age relationship for the ice core. Note that BC concentrations generally peak in summer to early autumn coincident with the boreal forest fire season.

6 Air Mass Back-Trajectory Modeling To identify likely source regions for BC, VA, S, and other chemical and elemental tracers measured in the ice cores, we used the National Oceanographic and Atmospheric Administration three-dimensional atmospheric trajectory model [Harris, 2004; Harris and Kahl, 1994]. The 1958 through 2002 atmospheric velocity fields from the ERA-40 reanalysis [Uppsala et al., 2005] were used to drive the model. Three-day trajectories were used because modeling [Flanner et al., 2007] suggests conversion in the atmosphere of hydrophobic to hydrophilic BC in ~1.5 days from which we estimate a lifetime of ~3 days. The 12-hourly trajectories were analyzed using the three-dimensional residence time analysis methodology described by Miller et al. [2002]. The lowest-bin residence time (Fig. S1) indicates transport within the lowest 1.5 km above ground (or sea) level.

7 Radiative Transfer Modeling BC surface radiative forcing calculations were conducted with the SNow, ICe, and Aerosol Radiative (SNICAR) model [Flanner et al., 2007]. Following Wiscombe and Warren [1980] and Warren and Wiscombe [1980], hemispheric radiative fluxes are derived in SNICAR from Mie ice and aerosol optical properties using multi-layer solution and delta-hemispheric mean approximation for multiple scattering [Toon et al., 1989]. We estimated surface incident solar radiative fluxes for the ice core site using the Shortwave NarrowBand (SWNB) radiative transfer model [Zender et al., 1997], applying typical sub-arctic summer clear-sky and cloudy atmospheric profiles. Because the D4 ice core site is located at high elevation in a region of little or no summer melting, we assumed an ice effective grain size of 100 µm. Snow aerosol radiative forcing increases with increasing grain size so our grain size choice contributes to a conservative forcing estimate. For BC optical properties, we used the mean of the measured median particle masses (1.82 femptograms) and geometric lognormal distribution (2.64) as well BC indices of refraction from Chang and Charalampopoulos [1990] and a particle density of 1250 kg m 3. In addition, we assumed that BC particles deposited in Greenland snow are sulfate-coated with a shell radius 1.33 times greater than the BC core, which is consistent with an industrial source. Together, these assumptions produce a BC mass absorption cross-section (MAC) of 8.5m 2 g 1 at 550 nm, greater than the recommended MAC of 7:5 m 2 g 1 for freshly-combusted BC [Bond and Bergstrom, 2006] but smaller than that of thicklycoated particles whose MAC can be amplified by a factor of 1.5 [Bond et al., 2006]. We generated a high-resolution lookup table of spectrally-resolved (470 band) albedo perturbation from varying amounts of BC and zenith angle for clear and cloudy skies. Using this table, we then estimated clear- and cloudy-sky radiative

8 forcing at half-hour resolution for the full 1788 to 2002 time series, using zenith angle computed from Berger [1978]. While clear- and cloudy-sky time series bound plausible forcing variability from sky conditions, there is surprisingly little reduction (~10%) in surface forcing under thickly-clouded skies over Greenland. This insensitivity occurs because (1) more of the insolation reduction caused by clouds is in the near-infrared, whereas the effect of BC on snow albedo is almost entirely in the visible; and (2) diffuse radiation has an effective zenith angle that is smaller than typical Greenland clear-sky conditions. Lowering the effective zenith angle increases forcing because radiation penetrates deeper in the snowpack. A monthly linear interpolation between clear sky and cloudy solar surface insolation from SWNB was derived to match that from reanalysis data at this location [Qian et al., 2006]. This weighting was applied to derive the all-sky forcing that we report (Fig. S3). Effective forcing is tropopause-adjusted radiative forcing scaled to represent the expected global mean temperature change from the forcing agent. Specifically, it is the response-to-forcing ratio of the agent relative to CO 2 [Hansen et al., 2005]. BC in snow has particularly high efficacy because it triggers snow-albedo feedback and grain coarsening. Flanner et al. [2007] found a BC/snow forcing efficacy of about 3.1. Instantaneous surface forcing estimates from SNICAR are scaled by 0:91 to represent tropopause-adjusted forcing, based on SWNB estimates spanning a wide range of sky conditions [Flanner et al., 2007] and then scaled by the efficacy factor (3.1). Effective forcing is likely a high estimate of the expected temperature change over non-melt regions of Greenland, where snow-albedo feedback is inhibited, but this forcing is appropriate for the Arctic as a whole [Flanner et al., 2007]. The largest source of uncertainty in our forcing estimates is from BC optical properties.

9 Fig. S3. Monthly BC concentration (A) and monthly averaged instantaneous and effective radiative forcing from black carbon measured in the D4 ice core over the period from 1870 to 1910.

10 Supplementary References Banta, J.R. and J.R. McConnell (2007), Annual accumulation over recent centuries at four sites in central Greenland, Journal of Geophysical Research, In Press. Berger, A. (1978), Long-term variations of daily insolation and quaternary climatic changes, J. Atmos. Sci., 35, Bond, T. C., and R. W. Bergstrom (2006), Light absorption by carbonaceous particles: An investigative review, Aerosol Sci. Technol.., 40(1), 27-67, doi: / Bond, T. C., G. Habib, and R. W. Bergstrom (2006), Limitations in the enhancement of visible light absorption due to mixing state, J. Geophys. Res., 111, D20211, doi: /2006jd Chang, H., and T. T. Charalampopoulos (1990), Determination of the wavelength dependence of refractive indices of _ame soot, Proc. Roy. Soc. London A, Math. and Phys. Sci., 430(1880), Chylek, P. B. Johnson, P. Damiano, K. Taylor and P. Clement (1995), Biomass Burning and Black Carbon in the GISP2 Ice Core, Geophys. Res. Lett., 22, 89-92, Dibb, J. E., R. W. Talbot, S. I. Whitlow, M. C. Shipham, J. R. Winterle, J. R. McConnell, and R. C. Bales (1996), Biomass burning signatures in the atmosphere and snow at Summit, Greenland: An event on 5 August 1994, Atmospheric Environment, V 30(4), p Flanner, M. G., C. S. Zender, J. T. Randerson, and P. J. Rasch (2007), Present day climate forcing and response from black carbon in snow, J. Geophys. Res., in press. Fine, P.M., Cass, G.R., and B.R.T. Simoneit (2004), Chemical characterization of fine particle emissions from the fireplace combustion of wood types grown in the Midwestern and western United States, Environmental Engineering Science, 21(3), Hansen, J., et al. (2005), Efficacy of climate forcings, J. Geophys. Res., 110, D18104, doi: /2005jd Harris, J.M. (2004), Summary Report Report 27, Climate Monitoring and Diagnostic Laboratories, National Oceanic and Atmospheric Administration, Boulder, CO, USA. Harris, J.M. and J.D. Kahl (1994), Analysis of 10-day isentropic flow patterns for Barrow, Alaska: J. Geophys. Res., 99, 25,845-25,855. Holdsworth, G., et al. (1996), Historical biomass burning: Late 19 th century pioneer agriculture recolution in northern hemisphere ice core data and its atmospheric interpretation, Journal of Geophysical Research, 101(D18), 23,371-23,334. McConnell, J. R., R. C. Bales, J. R. Winterle, H. Kuhns, and C. R. Stearns (1997), A lumped parameter model for the atmosphere-to-snow transfer function for hydrogen peroxide, Journal of Geophysical Research, 102(C12), McConnell, J.R., G.W. Lamorey, S.W. Lambert and K.C. Taylor (2002a), Continuous icecore chemical analyses using inductively coupled plasma mass spectrometry, Environ. Sci. Technol., 36(1), McConnell, J.R., G.W. Lamorey and M.A. Hutterli (2002b), A 250-year high-resolution record of Pb flux and crustal enrichment in central Greenland, Geophys. Res. Let., 29(23), 2130.

11 McConnell, J.R., A.J. Aristarain, J.R. Banta, P.R. Edwards, and J.C. Simões (2007), 20 th Century Doubling in Dust Archived in an Antarctic Peninsula Ice Core Parallels Climate Change and Desertification in South America, Proceedings of the National Academy of Sciences, 104(14), Miller, J.E., J.D.W. Kahl, F. Heller and J.M. Harris (2002), A three-dimensional residence-time analysis of potential summertime atmospheric transport to Summit, Greenland. Ann. Glaciol., 35, Qian, T., A. Dai, K. E. Trenberth, and K. W. Oleson (2006), Simulation of global land surface conditions from 1948 to Part I: Forcing data and evaluations, J. Hydrometeorology, 7, Schwarz, J.P., R.S. Gao, D.W. Fahey, D.S. Thomson, L.A. Watts, J.C. Wilson, J.M. Reeves, D.G. Baumgardner, G.L. Kok, S. Chung, M. Schulz, J. Hendricks, A. Lauer, B. Kärcher, J.G. Slowik, K.H. Rosenlof, T.L. Thompson, A.O. Langford, M. Lowenstein, K.C. Aikin (2006), Single-particle measurements of mid latitude black carbon and light-scattering aerosols from the boundary layer to the lower stratosphere, J. Geophys. Res. 111, D16207, doi: /2006jd Simoneit, B.R.T. (2002), Biomass burning a review of organic tracers for smoke from incomplete combustion, Applied Geochemistry, 17, Stephens, M., N. Turner and J. Sandberg (2003) Particle identification by laser induced incandescence in a solid state laser cavity. Appl. Optics, 42, Toon, O. B., C. P. McKay, T. P. Ackerman, and K. Santhanam (1989), Rapid calculation of radiative heating rates and photodissociation rates in inhomogeneous multiple scattering atmospheres, J. Geophys. Res., 94(D13), 16,287-16,301. Uppsala, S. M. and co-authors (2005), The ERA-40 re-analysis. QJRMS, 131 (612): Warren, S., and W. Wiscombe (1980), A model for the spectral albedo of snow. II: Snow containing atmospheric aerosols, J. Atmos. Sci., 37, Wilson, A.T. (1978), Pioneer agriculture explosion and CO 2 levels in the atmosphere, Nature, 273, Wiscombe, W. J., and S. G. Warren (1980), A model for the spectral albedo of snow. I: Pure snow, J. Atmos. Sci., 37, Zender, C. S., B. Bush, S. K. Pope, A. Bucholtz, W. D. Collins, J. T. Kiehl, F. P. J. Valero, and J. Vitko, Jr. (1997), Atmospheric absorption during the Atmospheric Radiation Measurement (ARM) Enhanced Shortwave Experiment (ARESE), J. Geophys. Res., 102(D25), 29,901-29,915.