• Adams, J. M., , and G. Piovesan, 2005: Long series relationships between global interannual CO2 increment and climate: Evidence for stability and change in role of the tropical and boreal-temperate zones. Chemosphere, 59 , 15951612.

    • Search Google Scholar
    • Export Citation
  • Baldocchi, D. D., , B. B. Hicks, , and T. P. Meyers, 1988: Measuring biosphere-atmosphere exchanges of biologically related gases with micrometeorological methods. Ecology, 69 , 13311340.

    • Search Google Scholar
    • Export Citation
  • Barford, C. C., and Coauthors, 2001: Factors controlling long- and short-term sequestration of atmospheric CO2 in a mid-latitude forest. Science, 294 , 16881691.

    • Search Google Scholar
    • Export Citation
  • Barnston, A. G., , and R. E. Livezey, 1987: Classification, seasonality, and persistence of low-frequency atmospheric circulation patterns. Mon. Wea. Rev., 115 , 10831126.

    • Search Google Scholar
    • Export Citation
  • Barr, A. G., , T. J. Griffis, , T. A. Black, , X. Lee, , R. M. Staebler, , J. D. Fuentes, , Z. Chen, , and K. Morgenstern, 2002: Comparing the carbon budgets of boreal and temperate deciduous forest stands. Can. J. For. Res., 32 , 813822.

    • Search Google Scholar
    • Export Citation
  • Black, T. A., and Coauthors, 2000: Increased carbon sequestration by a boreal deciduous forest in years with a warm spring. Geophys. Res. Lett., 27 , 12711274.

    • Search Google Scholar
    • Export Citation
  • Bousquet, P., , P. Peylin, , P. Ciais, , C. Le Quéré, , P. Friedlingtein, , and P. P. Tans, 2000: Regional changes in carbon dioxide fluxes of land and oceans since 1980. Science, 290 , 13421346.

    • Search Google Scholar
    • Export Citation
  • Braswell, B. H., , D. S. Schimel, , E. Linder, , and B. Moore III, 1997: The response of global terrestrial ecosystems to interannual temperature variability. Science, 278 , 870872.

    • Search Google Scholar
    • Export Citation
  • Buermann, W., , B. Anderson, , C. J. Tucker, , R. E. Dickinson, , W. Lucht, , C. S. Potter, , and R. B. Myneni, 2003: Interannual covariability in Northern Hemisphere air temperatures and greenness associated with El Niño-Southern Oscillation and the Arctic Oscillation. J. Geophys. Res., 108 .4396, doi:10.1029/2002JD002630.

    • Search Google Scholar
    • Export Citation
  • Cao, M., , and F. I. Woodward, 1998: Dynamic responses of terrestrial ecosystem carbon cycling to global climate change. Nature, 393 , 249252.

    • Search Google Scholar
    • Export Citation
  • Ciais, P., and Coauthors, 2005: Europe-wide reduction in primary productivity caused by the heat and drought in 2003. Nature, 437 , 529533.

    • Search Google Scholar
    • Export Citation
  • Cook, B. D., and Coauthors, 2004: Carbon exchange and venting anomalies in an upland deciduous forest in northern Wisconsin, USA. Agric. For. Meteor., 126 , 271295.

    • Search Google Scholar
    • Export Citation
  • Cramer, W., and Coauthors, 2001: Global response of terrestrial ecosystem structure and function to CO2 and climate change: Results from six dynamic global vegetation models. Global Change Biol., 7 , 357373.

    • Search Google Scholar
    • Export Citation
  • Dargaville, R. J., , S. C. Doney, , and I. Y. Fung, 2003: Inter-annual variability in the interhemispheric atmospheric CO2 gradient: Contributions from transport and the seasonal rectifier. Tellus, 55B , 711722.

    • Search Google Scholar
    • Export Citation
  • Deser, C., , and S. Phillips, 2006: Simulation of the 1976/77 climate transition over the North Pacific: Sensitivity to tropical forcing. J. Climate, 19 , 61706179.

    • Search Google Scholar
    • Export Citation
  • Dominguez, F., , and P. Kumar, 2005: Dominant modes of moisture flux anomalies over North America. J. Hydrometeor., 6 , 194209.

  • Dufresne, J-L., , P. Friedlingstein, , M. Berthelot, , L. Bopp, , P. Ciais, , L. Fairhead, , H. Le Treut, , and P. Monfray, 2002: On the magnitude of positive feedback between future climate change and the carbon cycle. Geophys. Res. Lett., 29 .1405, doi:10.1029/2001GL013777.

    • Search Google Scholar
    • Export Citation
  • Falge, E., and Coauthors, 2001: Gap filling strategies for defensible annual sums of net ecosystem exchange. Agric. For. Meteor., 107 , 4369.

    • Search Google Scholar
    • Export Citation
  • Favre, A., , and A. Gershunov, 2006: Extra-tropical cyclonic/anticyclonic activity in North-Eastern Pacific and air temperature extremes in Western North America. Climate Dyn., 26 , 617629.

    • Search Google Scholar
    • Export Citation
  • Flanagan, L. B., , L. A. Wever, , and P. J. Carlson, 2002: Seasonal and interannual variation in carbon dioxide exchange and carbon balance in a northern temperate grassland. Global Change Biol., 8 , 599615.

    • Search Google Scholar
    • Export Citation
  • Folland, C. K., and Coauthors, 2001: Observed climate variability and change. Climate Change 2001: The Scientific Basis, J. T. Houghton et al., Eds., Cambridge University Press, 99–182.

    • Search Google Scholar
    • Export Citation
  • Friedlingstein, P., and Coauthors, 2006: Climate–carbon cycle feedback analysis: Results from the C4MIP model intercomparison. J. Climate, 19 , 33373353.

    • Search Google Scholar
    • Export Citation
  • Fung, I. Y., , S. C. Doney, , K. Lindsay, , and J. John, 2005: Evolution of carbon sinks in a changing climate. Proc. Natl. Acad. Sci. USA, 102 , 1120111206.

    • Search Google Scholar
    • Export Citation
  • Goulden, M. L., , J. W. Munger, , S-M. Fan, , B. C. Daube, , and S. C. Wofsy, 1996: Measurements of carbon sequestration by long-term eddy covariance: Methods and a critical evaluation of accuracy. Global Change Biol., 2 , 169182.

    • Search Google Scholar
    • Export Citation
  • Goulden, M. L., , B. C. Daube, , S-M. Fan, , D. J. Sutton, , A. Bazzaz, , J. W. Munger, , and S. C. Wofsy, 1997: Physiological responses of a black spruce forest to weather. J. Geophys. Res., 102 , D24. 2898728996.

    • Search Google Scholar
    • Export Citation
  • Griffis, T. J., , T. A. Black, , K. Morgenstern, , A. G. Barr, , Z. Nesic, , G. B. Drewitt, , D. Gaumont-Guay, , and J. H. McCaughey, 2003: Ecophysiological controls on the carbon balances of three southern boreal forests. Agric. For. Meteor., 117 , 5371.

    • Search Google Scholar
    • Export Citation
  • Gu, L., and Coauthors, 2005: Objective threshold determination for nighttime eddy flux filtering. Agric. For. Meteor., 128 , 179197.

  • Gurney, K. R., and Coauthors, 2002: Towards robust regional estimates of CO2 sources and sinks using atmospheric transport models. Nature, 415 , 626630.

    • Search Google Scholar
    • Export Citation
  • Hember, R. A., , P. M. Lafleur, , and J. G. Cogley, 2005: Synoptic controls on summer evapotranspiration from a bog peatland in southern Canada. Int. J. Climatol., 25 , 793809.

    • Search Google Scholar
    • Export Citation
  • Hollinger, D. Y., , and A. D. Richardson, 2005: Uncertainty in eddy covariance measurements and its application to physiological models. Tree Physiol., 25 , 873885.

    • Search Google Scholar
    • Export Citation
  • Hollinger, D. Y., , S. M. Goltz, , E. A. Davidson, , J. T. Lee, , K. Tu, , and H. T. Valentine, 1999: Seasonal patterns and environmental controls of carbon dioxide and water vapour exchange in an ecotonal boreal forest. Global Change Biol., 5 , 891902.

    • Search Google Scholar
    • Export Citation
  • Hollinger, D. Y., and Coauthors, 2004: Spatial and temporal variability in forest–atmosphere CO2 exchange. Global Change Biol., 10 , 16891706.

    • Search Google Scholar
    • Export Citation
  • Houghton, R. A., 2003: Why are estimates of the terrestrial carbon balance so different? Global Change Biol., 9 , 500509.

  • House, J. I., , I. C. Prentice, , N. Ramankutty, , R. A. Houghton, , and M. Heimann, 2003: Reconciling apparent inconsistencies in estimates of terrestrial CO2 sources and sinks. Tellus, 55B , 345363.

    • Search Google Scholar
    • Export Citation
  • Humphreys, E. R., , T. A. Black, , K. Morgenstern, , Z. Li, , and Z. Nesic, 2005: Net ecosystem production of a Douglas-fir stand for 3 years following clearcut harvesting. Global Change Biol., 11 , 450464.

    • Search Google Scholar
    • Export Citation
  • Hurrell, D. A., , and H. van Loon, 1997: Decadal variations in climate associated with the North Atlantic Oscillation. Climatic Change, 36 , 301326.

    • Search Google Scholar
    • Export Citation
  • Janssens, I. A., and Coauthors, 2001: Productivity overshadows temperature in determining soil and ecosystem respiration across European forests. Global Change Biol., 7 , 269278.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77 , 437471.

  • Lafleur, P. M., , N. T. Roulet, , and S. W. Admiral, 2001: Annual cycle of CO2 exchange at a bog peatland. J. Geophys. Res., 106 , D3. 30713081.

    • Search Google Scholar
    • Export Citation
  • Lafleur, P. M., , N. T. Roulet, , J. L. Bubier, , S. Frolking, , and T. R. Moore, 2003: Interannual variability in the peatland-atmosphere carbon dioxide exchange at an ombrotrophic bog. Global Biogeochem. Cycles, 17 .1036, doi:10.1029/2002GB001983.

    • Search Google Scholar
    • Export Citation
  • Limpasuvan, V., , and D. L. Hartmann, 1999: Eddies and the annular modes of climate variability. Geophys. Res. Lett., 26 , 31333136.

  • Liu, J., , R. E. Stewart, , and K. K. Szeto, 2004: Moisture transport and other hydrometeorological features associated with the severe 2000/01 drought over the western and central Canadian Prairies. J. Climate, 17 , 305319.

    • Search Google Scholar
    • Export Citation
  • Massman, W. J., , and X. Lee, 2002: Eddy covariance flux corrections and uncertainties in long-term studies of carbon and energy exchanges. Agric. For. Meteor., 113 , 121144.

    • Search Google Scholar
    • Export Citation
  • Moncrieff, J. B., , Y. Malhi, , and R. Leuning, 1996: The propagation of errors in long-term measurements of land-atmosphere fluxes of carbon and water. Global Change Biol., 2 , 231240.

    • Search Google Scholar
    • Export Citation
  • Morgenstern, K., and Coauthors, 2004: Sensitivity and uncertainty of the carbon balance of a Pacific Northwest Douglas-fir forest during an El Niño/La Niña cycle. Agric. For. Meteor., 123 , 201219.

    • Search Google Scholar
    • Export Citation
  • Newman, M., , G. P. Compo, , and M. A. Alexander, 2003: ENSO-forced variability of the Pacific decadal oscillation. J. Climate, 16 , 38533857.

    • Search Google Scholar
    • Export Citation
  • North, G. R., , T. L. Bell, , R. F. Cahalan, , and F. J. Moeng, 1982: Sampling errors in the estimation of empirical orthogonal functions. Mon. Wea. Rev., 110 , 699706.

    • Search Google Scholar
    • Export Citation
  • Ogi, M., , K. Yamazaki, , and Y. Tachibana, 2004: The summertime annular mode in the Northern Hemisphere and its linkage to the winter mode. J. Geophys. Res., 109 .D20114, doi:10.1029/2004JD004514.

    • Search Google Scholar
    • Export Citation
  • Ogi, M., , K. Yamazaki, , and Y. Tachibana, 2005: The summer northern annular mode and abnormal summer weather in 2003. Geophys. Res. Lett., 32 .L04706, doi:10.1029/2004GL021528.

    • Search Google Scholar
    • Export Citation
  • Palmroth, S., , C. A. Maier, , H. R. McCarthy, , A. C. Oishi, , H-S. Kim, , K. H. Johnsen, , G. G. Katul, , and R. Oren, 2005: Contrasting responses to drought of forest floor CO2 efflux in a Loblolly pine plantation and a nearby Oak-Hickory forest. Global Change Biol., 11 , 421434.

    • Search Google Scholar
    • Export Citation
  • Patra, P. K., , M. Ishizawa, , S. Maksyutov, , T. Nakazawa, , and G. Inoue, 2005: Role of biomass burning and climate anomalies for land-atmosphere carbon fluxes based on inverse modeling of atmospheric CO2. Global Biogeochem. Cycles, 19 .GB3005, doi:10.1029/2004GB002258.

    • Search Google Scholar
    • Export Citation
  • Peylin, P., and Coauthors, 2005: Multiple constraints on regional CO2 flux variations over land and ocean. Global Biogeochem. Cycles, 19 .GB1011, doi:10.1029/2003GB002214.

    • Search Google Scholar
    • Export Citation
  • Ponton, S., , L. B. Flanagan, , K. P. Alstad, , B. G. Johnson, , K. Morgenstern, , N. Kljun, , T. A. Black, , and A. G. Barr, 2006: Comparison of ecosystem water-use efficiency among Douglas-fir forest, aspen forest, and grassland using eddy covariance and carbon isotope techniques. Global Change Biol., 12 , 294310.

    • Search Google Scholar
    • Export Citation
  • Press, W. H., , B. P. Flannery, , S. A. Teukolsky, , and W. T. Vetterling, 1986: Numerical Recipes: The Art of Scientific Computing. Cambridge University Press, 498 pp.

    • Search Google Scholar
    • Export Citation
  • Raich, J. W., , and W. H. Schlesinger, 1992: The global carbon dioxide flux in soil respiration and its relationship to vegetation and climate. Tellus, 44B , 8199.

    • Search Google Scholar
    • Export Citation
  • Richardson, A. D., , and D. Y. Hollinger, 2005: Statistical modeling of ecosystem respiration using eddy covariance data: Maximum likelihood parameter estimation, and Monte Carlo simulation of model and parameter uncertainty, applied to three simple models. Agric. For. Meteor., 131 , 191208.

    • Search Google Scholar
    • Export Citation
  • Richman, M. B., 1986: Rotation of principal components. Int. J. Climatol., 6 , 293335.

  • Rogers, J. C., 1997: North Atlantic storm track variability and its association to the North Atlantic Oscillation and climate variability of northern Europe. J. Climate, 10 , 16351647.

    • Search Google Scholar
    • Export Citation
  • Sacks, W. J., , D. S. Schimel, , R. K. Monson, , and B. H. Braswell, 2006: Model-data synthesis of diurnal and seasonal CO2 fluxes at Niwot Ridge, Colorado. Global Change Biol., 12 , 240259.

    • Search Google Scholar
    • Export Citation
  • Sakai, R. K., , D. R. Fitzjarrald, , and K. E. Moore, 2001: Importance of low-frequency contributions to eddy fluxes observed over rough surfaces. J. Appl. Meteor., 40 , 21782192.

    • Search Google Scholar
    • Export Citation
  • Schaefer, K., , A. S. Denning, , N. Suits, , J. Kaduk, , I. Baker, , S. Los, , and L. Prihodko, 2002: Effect of climate on interannual variability of terrestrial CO2 fluxes. Global Biogeochem. Cycles, 16 .1101, doi:10.1029/2002GB001928.

    • Search Google Scholar
    • Export Citation
  • Schmid, H. P., , C. S. Grimmond, , F. Cropley, , B. Offerle, , and H-B. Su, 2000: Measurements of CO2 and energy fluxes over a mixed hardwood forest in the midwestern United States. Agric. For. Meteor., 103 , 357374.

    • Search Google Scholar
    • Export Citation
  • Schmid, H. P., , H-B. Su, , C. S. Vogel, , and P. S. Curtis, 2003: Ecosystem-atmosphere exchange of carbon dioxide over a mixed hardwood forest in northern lower Michigan. J. Geophys. Res., 108 .4417, doi:10.1029/2002JD003011.

    • Search Google Scholar
    • Export Citation
  • Shindell, D. T., , G. A. Schmidt, , M. E. Mann, , D. Rind, , and A. Waple, 2001: Solar forcing of regional climate change during the Maunder Minimum. Science, 294 , 21492152.

    • Search Google Scholar
    • Export Citation
  • Stocker, T. F., and Coauthors, 2001: Physical climate processes and feedbacks. Climate Change 2001: The Scientific Basis, J. T. Houghton et al., Eds., Cambridge University Press, 417–470.

    • Search Google Scholar
    • Export Citation
  • Stoy, P. C., , G. G. Katul, , M. B. S. Siqueira, , J-Y. Juang, , K. A. Novick, , J. M. Uebelherr, , and R. Oren, 2006: An evaluation of models for partitioning eddy covariance-measured net ecosystem exchange into photosynthesis and respiration. Agric. For. Meteor., 141 , 218.

    • Search Google Scholar
    • Export Citation
  • Thompson, D. W. J., , and J. M. Wallace, 1998: The Arctic Oscillation signature in the wintertime geopotential height and temperature fields. Geophys. Res. Lett., 25 , 12971300.

    • Search Google Scholar
    • Export Citation
  • Thompson, D. W. J., , and J. M. Wallace, 2000: Annular modes in the extratropical circulation. Part I: Month-to-month variability. J. Climate, 13 , 10001016.

    • Search Google Scholar
    • Export Citation
  • Thompson, D. W. J., , and J. M. Wallace, 2001: Regional climate impacts of the Northern Hemisphere annular mode. Science, 293 , 8689.

  • Thompson, D. W. J., , J. M. Wallace, , and G. C. Hegerl, 2000: Annular modes in the extratropical circulation. Part II: Trends. J. Climate, 13 , 10181036.

    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., , G. W. Branstator, , D. Karoly, , A. Kumar, , N-C. Lau, , and C. Ropelewski, 1998: Progress during TOGA in understanding and modeling global teleconnections associated with tropical sea surface temperatures. J. Geophys. Res., 103 , C7. 1429114324.

    • Search Google Scholar
    • Export Citation
  • Twine, T. E., and Coauthors, 2000: Correcting eddy-covariance flux underestimates over a grassland. Agric. For. Meteor., 103 , 279300.

    • Search Google Scholar
    • Export Citation
  • von Storch, H., , and F. W. Zwiers, 1999: Statistical Analysis in Climate Research. Cambridge University Press, 484 pp.

  • Wallace, J. M., , and D. S. Gutzler, 1981: Teleconnections in the geopotential height field during the Northern Hemisphere winter. Mon. Wea. Rev., 109 , 784812.

    • Search Google Scholar
    • Export Citation
  • Wallace, J. M., , Y. Zhang, , and J. A. Renwick, 1995: Dynamic contribution to hemispheric mean temperature trends. Science, 270 , 780783.

    • Search Google Scholar
    • Export Citation
  • Wallace, J. M., , Y. Zhang, , and L. Bajuk, 1996: Interpretation of interdecadal trends in Northern Hemisphere surface air temperature. J. Climate, 9 , 249260.

    • Search Google Scholar
    • Export Citation
  • Wang, X. L., , H. Wan, , and V. R. Swail, 2006: Observed changes in cyclone activity in Canada and their relationships to major circulation regimes. J. Climate, 19 , 896915.

    • Search Google Scholar
    • Export Citation
  • Wilson, K., and Coauthors, 2002: Energy balance closure at FLUXNET sites. Agric. For. Meteor., 113 , 223243.

  • Wu, Q., , and D. M. Straus, 2004: AO, COWL, and observed climate trends. J. Climate, 17 , 21392156.

  • Zeng, N., , A. Mariotti, , and P. Wetzel, 2005a: Terrestrial mechanisms of interannual CO2 variability. Global Biogeochem. Cycles, 19 .GB1016, doi:10.1029/2004GB002273.

    • Search Google Scholar
    • Export Citation
  • Zeng, N., , H. Qian, , C. Roedenbeck, , and M. Heimann, 2005b: Impact of 1998–2002 midlatitude drought and warming on terrestrial ecosystem and the global carbon cycle. Geophys. Res. Lett., 32 .L22709, doi:10.1029/2005GL024607.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Locations of eddy covariance stations and time series of cumulative monthly NEP since January 1994. Monthly averages have been used as placeholders for periods of missing data.

  • View in gallery

    Experimental surface of uncertainty in measurements of ER [σ(δ)] based on the standard deviation of first-order differences in hourly nighttime elements, binned according to friction velocity u* and soil temperature Tsoil at the Mer Bleue bog (station SBBP).

  • View in gallery

    First rotated empirical orthogonal function for the (a) SLP pattern, (b) Z500 pattern, (c) standardized seasonal time series, and (d) monthly standard deviation for expansion coefficients. The boldface solid contour equals zero, thin solid contours indicate positive anomalies, and thin, broken contours indicate negative anomalies. SLP contour interval = 0.5 hPa. Z500 contour interval = 20 m.

  • View in gallery

    As in Fig. 3 but for the second rotated EOF.

  • View in gallery

    Composite maps illustrating the phases of (a), (b) SLP-1 during MAM; (c), (d) Z500-1 during MAM; (e), (f) SLP-1 during JJA; and (g), (h) Z500-1 during JJA. Filled contours indicate SAT composite anomalies (0.5°C intervals). Superimposed contours indicate composite anomalies of SLP (0.5-hPa intervals) or Z500 (20-m intervals).

  • View in gallery

    Composite maps illustrating the phases of (a), (b) SLP-2 during MAM; (c), (d) Z500-2 during MAM; (e), (f) SLP-2 during JJA; and (g), (h) Z500-2 during JJA. Filled contours indicate SAT composite anomalies (0.5°C intervals). Superimposed contours indicate composite anomalies of SLP (0.5-hPa intervals) or Z500 (20-m intervals).

  • View in gallery

    Correlation coefficients (×100) from regressing station variables onto expansion coefficients of (a) spring SLP-1, (b) spring Z500-1, (c) summer SLP-1, and (d) summer Z500-1. Upper-left box: SAT; upper-right box: P; lower-left box: GEP; lower-right box: ER. Boldface coefficients indicate significance at the 95% confidence level.

  • View in gallery

    As in Fig. 7, but for (a) spring SLP-2, (b) spring Z500-2, (c) summer SLP-2, and (d) summer Z500-2.

  • View in gallery

    Residual analysis of the (a), (b) relationships between spring SLP-2 and standardized CO2 flux residuals after controlling lag-1 autocorrelation; the (c), (d) relationships between spring SLP-1 and residuals from (a) and (b), and the (e), (f) relationships between spring P and residuals from (a) and (b).

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 3 3 0
PDF Downloads 1 1 0

Effects of Serial Dependence and Large-Scale Tropospheric Circulation on Midlatitude North American Terrestrial Carbon Dioxide Exchange

View More View Less
  • 1 Department of Forest Resources Management, University of British Columbia, Vancouver, British Columbia, Canada
  • | 2 Department of Geography, Trent University, Peterborough, Ontario, Canada
© Get Permissions
Full access

Abstract

Linear regression was used to relate modes of tropospheric circulation variability to estimates of gross ecosystem production (GEP) and ecosystem respiration (ER) measured at 14 midlatitude North American eddy covariance (EC) towers. The North Atlantic Oscillation (NAO) exhibited a north–south gradient in its effect on fluxes, with negative influence on fluxes at central and northeastern stations and positive influence on fluxes at southeastern stations. During spring, average values of GEP and ER within the northern “cold” sector decreased by 22 and 12 g C m−2 (18% and 11%), respectively, in response to a unit increase (+1 standard deviation) in the expansion coefficient of the NAO mode. Despite a northward advancement of the “warm” sector during summer, GEP and ER remained negatively correlated with the NAO at northern stations, decreasing on average by 48 and 30 g C m−2 (8% and 6%), respectively. During spring, the North Pacific Oscillation (NPO) reduced GEP and ER at central and northeastern stations on average by 20 and 7 g C m−2 (16% and 6%) and increased GEP and ER at southern and west coast stations on average by 53 and 49 g C m−2 (12% and 17%) in response to a unit increase in the NPO. This pattern persisted into summer, only shifted northward, with flux decreases of 19 and 24 g C m−2 (3% and 5%) at northern stations and increases of 72 and 82 g C m−2 (9% and 16%) at central stations. The direction of the flux response in each case was supported by synoptic conditions inferred from composite maps of North American circulation and gridded surface air temperature anomalies. The magnitude and timing of the relationships differed between stations and was attributed to differences in geographic location and plant functional type. Difficulty in the interpretation of significant correlations was attributed to the short sample length of typical EC records and unmodeled variability, including, for example, modulation by the NAO during high NPO. Despite these limitations, long-term monitoring EC stations show promise in characterizing the regional and ecosystem-specific carbon cycle response to low-frequency modes of tropospheric circulation variability and may play a critical role in validating ecosystem model responses to such phenomena.

Corresponding author address: Robbie A. Hember, Department of Forest Resources Management, University of British Columbia, Forest Sciences Centre, 2424 Main Mall, Vancouver, BC V6T 1Z4, Canada. Email: robbie@hember.name

Abstract

Linear regression was used to relate modes of tropospheric circulation variability to estimates of gross ecosystem production (GEP) and ecosystem respiration (ER) measured at 14 midlatitude North American eddy covariance (EC) towers. The North Atlantic Oscillation (NAO) exhibited a north–south gradient in its effect on fluxes, with negative influence on fluxes at central and northeastern stations and positive influence on fluxes at southeastern stations. During spring, average values of GEP and ER within the northern “cold” sector decreased by 22 and 12 g C m−2 (18% and 11%), respectively, in response to a unit increase (+1 standard deviation) in the expansion coefficient of the NAO mode. Despite a northward advancement of the “warm” sector during summer, GEP and ER remained negatively correlated with the NAO at northern stations, decreasing on average by 48 and 30 g C m−2 (8% and 6%), respectively. During spring, the North Pacific Oscillation (NPO) reduced GEP and ER at central and northeastern stations on average by 20 and 7 g C m−2 (16% and 6%) and increased GEP and ER at southern and west coast stations on average by 53 and 49 g C m−2 (12% and 17%) in response to a unit increase in the NPO. This pattern persisted into summer, only shifted northward, with flux decreases of 19 and 24 g C m−2 (3% and 5%) at northern stations and increases of 72 and 82 g C m−2 (9% and 16%) at central stations. The direction of the flux response in each case was supported by synoptic conditions inferred from composite maps of North American circulation and gridded surface air temperature anomalies. The magnitude and timing of the relationships differed between stations and was attributed to differences in geographic location and plant functional type. Difficulty in the interpretation of significant correlations was attributed to the short sample length of typical EC records and unmodeled variability, including, for example, modulation by the NAO during high NPO. Despite these limitations, long-term monitoring EC stations show promise in characterizing the regional and ecosystem-specific carbon cycle response to low-frequency modes of tropospheric circulation variability and may play a critical role in validating ecosystem model responses to such phenomena.

Corresponding author address: Robbie A. Hember, Department of Forest Resources Management, University of British Columbia, Forest Sciences Centre, 2424 Main Mall, Vancouver, BC V6T 1Z4, Canada. Email: robbie@hember.name

1. Introduction

The terrestrial northern midlatitudes currently act as a sink of carbon dioxide (CO2; Houghton 2003; House et al. 2003) with estimates ranging between 1.4 (Bousquet et al. 2000) and 2.4 Pg C yr−1 (Gurney et al. 2002) during the 1990s. These regions, therefore, contribute significantly to the global carbon budget, counteracting sources from fossil fuel burning and land use change. However, climate change is expected to have significant effects on net ecosystem production (NEP) of CO2 at northern latitudes (Cao and Woodward 1998; Cramer et al. 2001; Ciais et al. 2005). Given the vast quantities of carbon stored in the boreal, steppe, and temperate ecozones, understanding regional responses of NEP to future climate variability has practical relevance in forestry management, carbon accounting, and predicting the future atmospheric CO2 growth rate.

Recent inclusion of the carbon cycle into coupled global climate models suggests that future terrestrial carbon cycling may act as a positive feedback to anthropogenic warming in part through reductions in NEP (Fung et al. 2005; Friedlingstein et al. 2006). Although global model experiments provide valuable evidence of the long-term net impact of terrestrial cycling on atmospheric CO2, more studies are required to reduce the uncertainty in the contribution of individual regions (Dufresne et al. 2002). Climate-induced variability in regional NEP may be linked with low-frequency modes of tropospheric circulation variability. These modes consist of regimes, described by Stocker et al. (2001, p. 435) as local maxima in the probability density function of atmospheric fields, and abrupt shifts in those regimes, such as the Pacific decadal oscillation (PDO) or El Niño–Southern Oscillation (ENSO). Such modes may directly affect regional temperatures through advection, adiabatic expansion, and compression (Wallace et al. 1996), and affect regional hydrology through alteration of atmospheric moisture transport (Liu et al. 2004; Dominguez and Kumar 2005), storm-track trajectories (Rogers 1997; Favre and Gershunov 2006; Wang et al. 2006), and evaporation (Hember et al. 2005).

In the northern midlatitudes, regional carbon cycle studies have suggested that the influence of alternating climate anomalies associated with standing atmospheric planetary waves has a tendency to cancel their contribution to the global carbon budget (Schaefer et al. 2002; Zeng et al. 2005a). Other studies have noted the importance of regional anomalies, such as the European heat wave of 2003 (Ciais et al. 2005) and the extended drought throughout subtropical latitudes during 1998–2002 (Zeng et al. 2005b).

Local-scale studies using eddy covariance (EC) measurements of NEP (Baldocchi et al. 1988) have also recognized the influence of climate anomalies associated with large-scale dynamical forcing. For example, warm temperatures following the 1997/98 ENSO event coincided with earlier leaf emergence and enhanced NEP at a southern boreal deciduous forest in Saskatchewan, Canada (Black et al. 2000). On the other hand, Morgenstern et al. (2004) showed that warmer temperatures during an ENSO event reduced NEP at a coastal Douglas fir forest in British Columbia, Canada. Previously this type of analysis has been limited by the sample size of typical EC records. However, data records at many stations now exceed or are approaching decennial time spans. As these time series lengthen, they may provide independent estimates of carbon cycle responses to climate trends and extreme anomalies, as they provide accurate indicators of temporal variability and can distinguish the individual biological components, gross ecosystem production (GEP) and ecosystem respiration (ER; Goulden et al. 1996; Hollinger et al. 2004; Morgenstern et al. 2004).

In this study, we investigated connections between contemporary large-scale climate variability and CO2 fluxes throughout a range of midlatitude North American ecosystems. To achieve this, we used linear regression to identify the presence of significant relationships between estimates of GEP and ER at 14 EC stations and modes of Northern Hemisphere tropospheric circulation variability, identified herein as the North Atlantic Oscillation (NAO) and the North Pacific Oscillation (NPO). Although such relationships may be prevalent over a range of temporal scales, we limited the current study to the analysis of seasonal connections by integrating CO2 fluxes over spring [March–May (MAM)] and summer [June–August (JJA)], when ecosystems were most productive.

Climate signals in the seasonal time series of GEP and ER may be obscured by the presence of serial dependence as a result of internal feedback mechanisms. Studies of atmospheric CO2 growth rate, inversion models, and vegetation indices show evidence that the effects of climate anomalies cascade through the biosphere (Braswell et al. 1997; Dargaville et al. 2003; Adams and Piovesan 2005; Patra et al. 2005; Peylin et al. 2005). While perturbations in shortwave radiation may immediately affect plant growth, the response in heterotrophic respiration may exhibit considerable lag time (Raich and Schlesinger 1992). In boreal and temperate ecozones, Braswell et al. (1997) showed that annual values of normalized difference vegetation index (NDVI) were correlated with immediate temperatures and inversely correlated with antecedent temperatures. Although serial correlation may be explained by changes in soil nutrient and water availability, in many cases correlations are weak, and possible spatial dependence within biome samples makes interpretation difficult. The presence of temporal dependence in ecosystem CO2 fluxes has also been difficult to demonstrate using EC measurements. At Harvard Forest, Barford et al. (2001) showed that late summer (August–September) NEP was positively correlated with ER of the previous winter. It may be argued that the insulation of soils induced by increased winter snow cover (or earlier onset of snow cover) may have persisted in its effect on ER into summer, which also corresponds with enhanced late summer NEP through increased nutrient availability. Using the same analysis procedure, Hollinger et al. (2004) showed that Howland Forest exhibited positive serial autocorrelation between winter and early spring ER and negative serial autocorrelation between winter and late fall (October–November) ER, while no such dependence between ER and NEP was found.

To account for the potential importance of serial dependence in detecting climate signals, we quantified lag-1 autocorrelation in seasonal GEP and ER and retested mode–CO2 flux relationships using partial correlation analysis to control for seasonal dependence in CO2 fluxes. Although these linear models likely oversimplify the actual relationships, small CO2 flux samples prohibit the use of more sophisticated nonlinear models. To address this limitation, we discuss potential functions that underlie the main relationships and the degree to which interdependencies between modes may contribute to nonstationary behavior in the individual mode–CO2 flux relationships.

2. Data and methods

a. Station descriptions

Continuous half-hourly measurements of CO2 flux and ancillary environmental variables were provided by 14 long-term monitoring FLUXNET stations. Data were collected from online data information systems maintained by Ameriflux (http://public.ornl.gov/ameriflux/data-access-select.shtml) and Fluxnet Canada (http://fluxnet.ccrp.ec.gc.ca), depending on station affiliation (see Table 1). Figure 1 shows the location and cumulative monthly time series of NEP calculated for each station. The sample is representative of mature communities, including five deciduous forests, seven coniferous forests, one grassland, and one peatland. The sample was chosen largely based on the length, continuity, and quality of EC measurements. Measurements were processed (i.e., cleaned but not gap-filled) by investigators at each station prior to retrieval, including coordinate rotation of the wind components, correction for variation in air density, correction for loss of high-frequency flux, compensation for changes in canopy storage of CO2, and removal of erroneous measurements resulting from sensor malfunction and periods of interference from the measurement infrastructure. Flux time series varied between stations and some records consisted of extended periods of missing data. Samples used in the analysis were arrived at by omitting long periods of missing data, defined as data gaps that exceeded 1 month, and gap-filling short periods of missing data, as outlined in section 2b. Thus, the analysis consisted of discontinuous seasonal time series based on the availability of data at each station, spanning 1994–2006.

b. CO2 flux data and treatment

A common gap-filling procedure was applied to each station record to minimize inconsistencies and produce estimates of the uncertainty in seasonal CO2 fluxes due to the gap-filling procedure. Missing hourly values were filled using empirical relationships between CO2 fluxes and ancillary meteorological variables (Falge et al. 2001). Measurements of NEP were derived from the summation of the turbulent and storage terms (NEP = Fc + Fs). Many studies have shown that the EC technique underestimates NEP during calm conditions fairly consistently in different ecosystems (Hollinger et al. 1999; Barford et al. 2001; Lafleur et al. 2001; Barr et al. 2002; Flanagan et al. 2002; Morgenstern et al. 2004; Gu et al. 2005). The dependence between nighttime NEP and friction velocity (u*) was tested using a method developed by Gu et al. (2005). Threshold values (u*th), marking the level below which NEP measurements became dependent on u*, closely reflected published values. Therefore, measurements of NEP were removed for periods when u* < u*th following the previously reported values of u*th listed in Table 1 prior to the gap-filling procedure.

ER was set equivalent to NEP during the inactive growth period (i.e., PAR ≤ 0, Tair ≤ −2, or Tsoil ≤ 0), where PAR is photosynthetically active radiation (μmol m−2 s−1), Tair is the air temperature at the top of the canopy (°C), and Tsoil is the soil temperature at 5 cm below the surface (°C). Initial estimates of ecosystem respiration, ERmod, were fitted to Tsoil using a “Q10” model:
i1520-0442-21-4-751-e1
where Rref is the expected efflux of CO2 at Tref (10°C), Q10 is the factor by which ER increases for a 10°C increase in temperature, and ε is the residual error. For the purpose of understanding how best to scale measurement uncertainty in the Q10 model, errors in ER were investigated after removing the seasonal and diurnal cycles. Error in hourly measurements generally increased with temperature and decreased with turbulent mixing, as illustrated by an experimental surface of the uncertainty σ(δ) calculated for a range of bin-averaged Tsoil and u* at the south boreal bog peatland (SBBP) station (Fig. 2). Here, elements of σ(δ) were calculated as the standard deviation (sd) of the absolute values of first-order differences in ER anomalies. That is, differences between hourly elements of normalized ER were assumed to provide a reasonable proxy for more sophisticated estimates of the random error associated with instrument malfunction, footprint heterogeneity, etc. The dispersion in σ(δ) was similar across the stations to that reported for the east boreal spruce hemlock (EBSH) station using multiple measurement systems with a ceiling of ∼3 μmol m−2 s−1 (Richardson and Hollinger 2005; Hollinger and Richardson 2005). Maximum likelihood estimates of the Q10 model parameters were found by minimizing the scaled absolute deviation criterion (Press et al. 1986; Richardson and Hollinger 2005):
i1520-0442-21-4-751-e2
where χ2 is the likelihood term and σ is the model uncertainty, approximated from experimental surfaces of σ(δ), as illustrated in Fig. 2. As in Humphreys et al. (2005), seasonal dynamics in ER were represented by adjusting Rref along moving time segments (or windows) to remove differences between measured ER and ERmod. The sample size of each window was set to 48 h, while the window increment was set to 10 h.

Increased uncertainty in ER toward low u* in Fig. 2 may be related to the so-called carbon counting problem. In general, negative anomalies in hourly ER occur simultaneously with extremely low u* and during nights with persistent thermodynamic stability. This apparent bias in EC measurements seems to indicate failure of EC systems to capture the full range of variability in Fs, perhaps as a result of accumulation of CO2 near (or within) the soil surface. While positive anomalies are commonly observed following resumed turbulence during subsequent hours, it is unclear to what degree this compensates for previous underestimation. As a result, it is argued that splicing ERmod estimates into the discontinuous time series of ER to replace missing values may introduce unnecessary uncertainty. To avoid this, we chose to use ERmod explicitly in subsequent analyses. However, because the windowing technique removes differences between measured ER and ERmod on time scales exceeding 2 days, this decision has only a small influence on seasonal estimates of ER used here.

Assimilation of CO2 during photosynthesis was approximated by GEP = NEP + ER during periods of active growth, and by GEP = 0 during periods of inactive growth. Initial estimates of GEP were filtered for outliers (±3 sd) on a seasonal basis. Gaps in GEP during the active growth period were filled using a light-response function applied during summer:
i1520-0442-21-4-751-e3
where α and Amax express the canopy-scale apparent quantum yield (mol CO2 mol−1 photon) and light-saturated photosynthetic capacity, respectively. Once again, parameters were fitted according to maximum likelihood estimation—this time scaling errors according to σ(ε) derived from polynomial fits to PAR. The initial model estimates were then adjusted using the same windowing technique described above for modeled ER. For rare gaps in GEP that exceeded 48 h, missing values were filled using the mean diurnal variation during days prior to and following each gap.

Potential errors in EC-based CO2 flux estimates arise from instrument calibration and microclimatic effects, including inadequate representation of CO2 advection, low-frequency turbulent fluxes, storage within the canopy (Massman and Lee 2002; Moncrieff et al. 1996; Sakai et al. 2001; Twine et al. 2000), and lack of energy balance closure (Wilson et al. 2002; Schmid et al. 2003). Error may also arise from the selection of u*th (Barford et al. 2001; Barr et al. 2002; Morgenstern et al. 2004), extrapolation of the Q10 model to daytime ER (Janssens et al. 2001), and the windowing adjustment of Rref (outlined above). While these errors introduce considerable uncertainty in the absolute magnitude of CO2 fluxes, there is little evidence available to indicate that these mechanisms strongly contribute to nonsystematic errors. Uncertainty in seasonal estimates of GEP and ER was briefly investigated by summing the compounded random error associated with hourly measurements (Morgenstern et al. 2004; Humphreys et al. 2005) and uncertainty associated with the gap-filling procedure. Errors in the model parameters [Eqs. (1) and (3)] were found by bootstrap resampling (von Storch and Zwiers 1999), drawing on 500 random samples from the dataset to calculate 95% confidence intervals for ER and missing GEP. Using this method, estimates of the average uncertainty were below 1 sd of seasonal values of GEP and ER, which suggests that the temporal variability of EC-based measurements is generally robust.

Figure 1 shows the cumulative monthly time series of NEP resulting from filling small gaps using the procedure outlined above. All 14 ecosystems were net sinks of atmospheric CO2 during the study period. Using monthly averages in place of missing months, the long-term (1994–2005) accumulation of carbon from these ecosystems ranged between 260 and 6500 g C m−2. Multiannual variability in NEP was, in most cases, only subtly noticeable. However, it is apparent from Fig. 1 that individual years can influence carbon budgets on decadal time scales. For example, productive years at the north temperate tallgrass prairie (NTTP) and north temperate ash maple (NTAM) stations during 2002 and 2003, respectively, counteracted multiple previous unproductive years. Long-term trends are also noticeable at the east temperate oak maple (ETOM), north boreal black spruce (NBBS), and south boreal aspen hazel (SBAH) stations. On average, spring GEP and ER across stations were 225 and 185 g C m−2, while the average standard deviation in spring GEP and ER across stations was 53 and 38 g C m−2. Similarly, average summer GEP and ER were 698 and 507 g C m−2 with average standard deviations of 76 and 59 g C m−2.

c. Atmospheric variables

Monthly-mean values of sea level pressure (SLP) and 500-hPa geopotential height (Z500) were extracted from the National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis project (Kalnay et al. 1996) provided by the National Oceanic and Atmospheric Administration (NOAA)/Office of Oceanic and Atmospheric Research (OAR)/Earth System Research Laboratory (ESRL)/Physical Sciences Division (PSD), Boulder, Colorado (http://www.cdc.noaa.gov/). Northern Hemisphere (20°–90°N) anomalies of SLP and Z500, spanning 1948–2006, were then used to derive modes of tropospheric circulation variability, as described in section 3. Large-scale fields of surface air temperature (SAT) and precipitation (P) were also collected to help interpret results. Monthly-mean SAT was extracted from the NCEP–NCAR reanalysis and used in composite analysis to further understand the synoptic effects of the modes on North American surface climate. Monthly total P was extracted from the North American Regional Reanalysis (NARR), also provided by the NOAA/OAR/ESRL/PSD. The NARR dataset provided strong agreement with most gauge measurements of monthly total precipitation at the EC stations and was, therefore, used to gap-fill missing P. Table 2 lists statistics for the linear regression between station estimates and NARR values corresponding to the nearest grid cell (∼30-km resolution). Consistent dampening of variability in NARR estimates can be explained by the difference in spatial scales. NARR estimates used to fill missing station measurements were, therefore, scaled by the regression coefficients to remove biases.

3. EOF analysis

Modes of tropospheric circulation variability are most often derived from teleconnection indices (Wallace and Gutzler 1981) or seasonally varying EOF analyses (Barnston and Livezey 1987). To avoid interpretive difficulties and sampling errors associated with calculation of separate seasonal EOF solutions, we derived the modes from EOF analyses of winter [December–March (DJFM)] monthly SLP and Z500 anomalies and constructed continuous monthly time series for each mode by regressing monthly SLP and Z500 anomalies onto their respective DJFM EOF patterns. Prior to calculation of the EOFs, monthly SLP and Z500 anomalies were weighted by the square root of the cosine of the latitude to remove variation in grid cell area. Coordinate transformations were applied to the first 10 EOFs in EOF-state space using varimax rotation (Richman 1986). For comparison with CO2 fluxes, we chose to limit the analysis to the first (SLP-1 and Z500-1) and second (SLP-2 and Z500-2) modes of variability.

The resulting time series of expansion coefficients (PCs) reflected the behavior of the dominant cold-season modes of Northern Hemisphere variability and provided a convenient means of investigating their relevance during spring and summer. Figures 3 and 4 show the DJFM EOF patterns and recent seasonal time series of the first and second EOFs, respectively. Figures 3d and 4d show the seasonal cycle in the variation of the PCs, with a tendency toward reduced circulation variability during summer. The Z500 modes closely resembled those calculated by performing separate EOF analyses for each season (Barnston and Livezey 1987). They showed that Z500-1 was generally a permanent feature of tropospheric circulation, while Z500-2 was less prevalent during the warm season.

a. The NAO

Prior to rotation, the first EOFs (SLP-1 and Z500-1) resembled the Arctic Oscillation (AO) in the lower troposphere (Thompson and Wallace 1998, 2000; Thompson et al. 2000) and the Northern Annular Mode (NAM) in the middle troposphere (Limpasuvan and Hartmann 1999; Ogi et al. 2004). Rotation accentuated the main centers of action and more closely resembled the NAO (Barnston and Livezey 1987; Wallace et al. 1996; Hurrell and van Loon 1997). The first two rotated EOFs each accounted for 15% of the total variance in both SLP and Z500 anomalies and were sufficiently separated from subsequent eigenvalues according to estimates of the error developed by North et al. (1982). On a seasonal time scale, SLP-1 and Z500-1 were strongly correlated during DJF (r = 0.95) and MAM (r = 0.81), but poorly correlated during JJA (r = 0.50) owing to reduced variability in SLP-1 (Fig. 3d). There was no significant autocorrelation in spring or summer PCs of SLP-1 or Z500-1.

Composite maps of circulation and SAT anomalies during SLP-1 and Z500-1 are shown in Fig. 5. During spring, the positive phase of the first EOFs was associated with intensified meridional pressure gradient over eastern North America and SAT anomalies of −2° and +1°C in northeastern and southeastern North America, respectively (Fig. 5a), while the negative phase was reciprocal (Fig. 5b). During summer, the centers of action associated with the spring NAO remained intact but were shifted north and with weakened wave structure across the North American sector (Figs. 5e,f). The effect of SLP-1 on summer SAT anomalies was consistent with that during spring, shown here, and during winter, as outlined by Thompson and Wallace (2001). During periods of intensified meridional pressure gradient (high SLP-1), polar air is contained by the polar vortex, allowing tropical systems to advance north to produce positive SAT anomalies in central Canada (Fig. 5e). During periods of weakened meridional pressure gradient (low SLP-1), polar systems advance south (Fig. 5f). These results for summer are consistent with those of Ogi et al. (2004, 2005, who attributed positive SAT anomalies in northern midlatitude continental regions with the formation of an upper-level double jet and associated blocking events.

In contrast, the warm-season configuration of Z500-1 (Figs. 5g,h) differed markedly from that of the AO or NAM. During the positive phase of Z500-1, eastern North America tended toward negative SAT anomalies. During the negative phase, positive SAT anomalies tended to span the Boreal Shield and strong negative SAT anomalies occurred over the west coast of North America. This is similar in structure, but opposite in direction, to that described by Shindell et al. (2001) in retrospective model simulations of the annual relationship between the AO and SAT in the North American sector.

b. The NPO

The dominant center of action in the second mode (SLP-2 and Z500-2) was situated in the central Pacific sector (40°N, 160°W), with the positive phase marked by intensification of the low pressure (Figs. 4a,b). After rotation, SLP-2 and Z500-2 accounted for 11% and 13% of the total variance and were also free from sampling errors (North et al. 1982). Seasonal time series of SLP-2 and Z500-2 were moderately correlated during spring (r = 0.86), while in summer, extreme positive values of Z500-2 occurred across the full range of values in SLP-2, resulting in poor correlation (r = 0.54). There was no seasonal autocorrelation in spring or summer values of SLP-2 or Z500-2.

The pattern of SLP-2 resembled the cold ocean–warm land (COWL) mode (Wallace et al. 1995, 1996; Wu and Straus 2004), while Z500-2 resembled the Pacific–North American (PNA) teleconnection, described by Barnston and Livezey (1987). These modes were not strongly correlated with indices of tropical variability on seasonal or annual time scales. However, Fig. 4c shows that extratropical Pacific circulation variability, particularly in SLP, was affected by the 1998 ENSO event. On decadal time scales, low-frequency variability in SLP-2 and Z500-2 resemble that of the PDO (with tropospheric modes leading). Thus, SLP-2 and Z500-2 appear to integrate the combined effects of variability in tropical convection, memory in sea surface temperature, teleconnectivity in general circulation of the atmosphere, and stochastic variability (Newman et al. 2003; Trenberth et al. 1998; Deser and Phillips 2006). In accordance with Folland et al. (2001, p. 154), these modes are identified herein as the NPO.

Composite maps of circulation and SAT anomalies for SLP-2 and Z500-2 were closely associated with the wave structure over North America (Fig. 6). During the positive phase of SLP-2, western coastal regions exhibited positive SAT anomalies, while negative SAT anomalies presided over interior regions. The negative phase resulted in reciprocal effects. Composites for Z500-2 closely reflected the trough–ridge–trough configuration commonly associated with the PNA teleconnection pattern.

4. Regression analysis

The degree of serial dependence and long-term trends in spring and summer CO2 fluxes is shown in Table 3. Spring fluxes were regressed against fluxes from the preceding winter (DJF), while summer fluxes were regressed against fluxes from the preceding spring. Winter GEP was nonzero only at the west temperate Douglas fir (WTDF) station. At all other stations, spring GEP was therefore regressed against winter ER [see Barford et al. (2001) and Hollinger et al. (2004) for discussion of possible lag effects between winter ER and GEP]. Significant lag-1 autocorrelation was detected at five of the stations and linear-in-time trend was detected at four stations. Linear regression analysis of the mode–CO2 flux relationships is shown in Fig. 7 (NAO) and Fig. 8 (NPO). Values at each station express the Pearson correlation coefficients (×100) between each respective mode and station values of SAT (upper-left boxes), P (upper-right boxes), GEP (lower-left boxes), and ER (lower-right boxes). Bold coefficients indicate statistical significance at the 95% confidence level. Partial cross correlations between fluxes and the modes are indicated in Tables 4 and 5.

a. NAO–CO2 flux relationships

Using an ecosystem model, Schaefer et al. (2002) found no relationship between the AO and North American CO2 flux variability during 1983–93. On the other hand, Buermann et al. (2003) showed a weak inverse correlation between spring AO and NDVI throughout parts of east-central North America during 1982–98. In this study, correlation between the NAO and midlatitude fluxes was modest and showed similar spatial structure to those previously found based on the AO index. During spring, both SAT and P were inversely correlated with the NAO across several of the northern stations. Correlations with GEP and ER were also consistently negative across the northern stations, but this effect was not significant at the 95% confidence level. Correlation in the southeast stations was positive, in agreement with SAT composites (Fig. 4). During summer, the NAO showed evidence of reversing its impact on SAT anomalies across the northern stations (Fig. 5c), which is in agreement with Ogi et al. (2005), but this did not have a significant effect on fluxes.

Explicit representation of serial dependence generally improved the relationships between the NAO and fluxes, albeit during different seasons, and often only in GEP or ER. During spring, effects of the NAO became significant at EBSH, NBBS, and the south temperate maple tulip (STMT) station. The direction of the residual flux response to the NAO was consistent with the composite maps and the analysis of Buermann et al. (2003); fluxes decreased in the northeast and increased in the southeast. Average spring residual values of GEP and ER within the northern “cold” sector [EBSH, NBBS, NTAM, NTTP, SBAH, SBBP, south boreal black spruce (SBBS), and south boreal jack pine (SBJP) stations] decreased by 22 and 12 g C m−2 (18% and 11% of the seasonal mean), respectively, in response to a unit increase (+1 sd) in the expansion coefficient of the NAO mode. During summer, effects of the NAO on residual values of GEP were significant at the east temperate oak hickory (ETOH), east temperate pine plantation (ETPP), NTAM, SBBP, and WTDF stations. It is clear from the composite maps that SLP-1 and Z500-1 reflect different phenomenon and differ in their influence over North American synoptic conditions during summer. That is, the positive phase of the summer AO (SLP-1) and the negative phase of the summer NAO (Z500-1) increased SAT anomalies throughout large areas of North America. In addition, SLP-1 and Z500-1 lead to important differences in pressure and SAT over the southeastern United States, which may help to explain their contrasting effects on GEP at ETOH and ETPP (Table 4). The strong inverse effects of both SLP-1 and Z500-1 on summer GEP at NTAM were not consistent with SAT anomalies. Summer GEP was positively correlated with Z500-1 WTDF, in agreement with negative height anomalies over the eastern extratropical Pacific basin. Thus, despite a northward advancement of the “warm” sector during summer, residual values of GEP and ER generally remained negatively correlated with the NAO at central and northern stations (i.e., NTAM, SBAH, SBBP, SBBP, SBBS, SBJP), decreasing on average by 48 and 30 g C m−2 (8% and 6%), respectively.

As discussed earlier, Barford et al. (2001) found evidence to suggest that late summer NEP was positively correlated with rates of decomposition during the preceding winter at the ETOM station. In the current study, the seasonal time series of winter ER and summer NEP at ETOM exhibited increasing linear trends, making it difficult to verify the significance of serial dependence using simple linear regression analysis. Simply detrending winter and spring ER anomalies by the linear-in-time fit prior to calculating the lag-1 serial autocorrelation at ETOM decreased the initial correlation to within the probability-of-chance occurrence (i.e., p = 0.04 changed to p = 0.37). This caveat applies to all results of the current study as we made no attempt to investigate the consequences of weak nonstationary behavior in the time series of fluxes at ETPP, a juvenile plantation, ETOM, and SBAH (Table 3). Such behavior at ETOM may help explain why representation of serial dependence in spring ER did not improve the relationship with the NAO.

b. NPO–CO2 flux relationships

Station fluxes in the north-central and eastern stations were inversely correlated with the NPO during spring (Fig. 8). This relationship was strongest between SLP-2 and fluxes at the north-central and northeast stations and was consistent with a negative (positive) correlation with SAT (P). Spring Z500-2 showed a similar behavior, but it only significantly affected fluxes at NBBS. During summer, the effect of the NPO persisted at NBBS and SBBS and expanded south, where fluxes were positively correlated with SLP-2 at NTAM, STMT, and ETOM, and with Z500-2 at NTTP and NTAM.

Variability in Z500-2 had a pronounced effect on SAT anomalies during spring and summer (Fig. 6). However, this effect only extended to fluxes at NTTP during summer (Fig. 8d). Controlling for serial dependence resulted in improvement of the correlation between Z500-2 and spring flux residuals at WTDF, while correlation with summer fluxes at NTTP remained strong (Table 5). Composites in Figs. 6g and 6h and weak positive correlations with SAT at NTTP and WTDF suggest that the positive Z500-2 controlled the intensity of warm-air advection into western regions during both spring and summer. Differences in the response timing at WTDF and NTTP can be attributed to differences in ecosystem type and offsets in the start of the growing season at each station. That is, the NPO may affect spring carbon cycling in coastal temperate ecosystems, such as WTDF, because spring temperature modulation occurs within the range necessary for growth, whereas leafout in continental regions (e.g., NTTP) has not yet begun (Flanagan et al. 2002). The peak growing period at NTTP was June and July; thus, NPO correlations were strongest at that time. These results indicate that, while individual modes may control fluxes over large geographic regions, they may exhibit significantly different response times and directions, in accordance with characteristics of the local climate and ecosystem type. The negative correlation between SLP-2 and spring fluxes across the northern stations was generally supported by partial correlation analysis, whereas the correlation between both modes and GEP at ETPP became insignificant, perhaps as a result of a long-term trend (Table 3). On average, residual values of spring GEP and ER decreased at central and northern stations by 20 and 7 g C m−2 (16% and 6%) and increased at southern and west coast stations by 53 and 49 g C m−2 (12% and 17%), respectively, in response to a unit increase in the NPO. With the exception of west alpine fir spruce (WAFS), whose summer flux residual was inversely related to the NPO, the positive correlation between the NPO and summer flux residuals improved or remained strong at southern stations (e.g., ETOM, NTTP, NTAM, STMT). The residual values of summer GEP and ER decreased at northern stations by 19 and 24 g C m−2 (3% and 5%), while central stations (NTAM, NTTP, STMT) exhibited average increases in GEP and ER of 72 and 82 g C m−2 (9% and 16%) in response to a unit increase in the NPO.

c. Lag effects

Several stations were inversely correlated with modes of the preceding season. Tables 6 and 7 show slope coefficients from linear regression analysis of seasonal fluxes with lag-1 values of the first and second modes, respectively. For example, fluxes at two western stations were correlated with the preceding value of the NAO. During spring, WTDF fluxes and SAT were inversely correlated with both winter values of SLP-1 and Z500-1, whereas summer fluxes and P were positively correlated with spring values of SLP-1 and Z500-1 at NTTP. While variability in spring SAT responded similarly between the two sites, we suspect that differences in the influence over fluxes arose once again because of overall differences in average spring temperature. During summer, influence of preceding NAO was reduced at WTDF. The contrasting response between the negative winter NAO–spring flux relationship at WTDF and the positive spring NAO–summer flux relationship observed at NTTP appears to be caused by a positive correlation in P. GEP and ER were also negatively correlated with the preceding value of the NAO. In each case, correlations with P and SAT were opposite. Thus, it is difficult to determine if these relationships were exclusively driven by variability in P, or if differences in temperature sensitivity and water use efficiency associated with semiarid grasslands versus the characteristics of humid boreal or coastal ecosystems (Ponton et al. 2006) plays a role.

5. Discussion

In spite of the relatively simple analysis used here (i.e., linear regression), we found several EC-based records of GEP and ER were significantly correlated with the NAO and the NPO. Relationships corresponded well with changes in synoptic conditions inferred from planetary wave structure and SAT anomalies in composite analysis maps. There was reasonably strong spatial coherence in correlations, given the degree of complexity expected in flux records due to local processes, ecophysiological differences, and measurement errors. Negative correlations in the northeast (EBSH and NBBS) and positive correlation in the southeast (STMT) during spring were consistent with the previously described influence of the AO/NAO pattern on the occurrence of cold-air outbreaks throughout southeastern regions of North America (Thompson and Wallace 2001; Buermann et al. 2003). During summer, SLP-1 (the AO) continued to inhibit fluxes at three stations, despite a reverse influence over SAT. The midtropospheric signature of the NAO was coupled with heights in the northeast Pacific basin, which caused a strong east–west influence over SAT anomalies and correlation with fluxes at WTDF and NTAM. The NPO exhibited widespread influence over fluxes during spring and summer.

Explicit representation of serial dependence in seasonal GEP and ER elucidated or improved strong correlations. Improvement of the mode–flux relationship (e.g., between the NAO and spring GEP at EBSH) may indicate that biogeochemical feedback mechanisms existed in flux records despite failing to detect serial dependence in simple linear regression analyses.

Scaling slope coefficients by the average spring CO2 flux at each station suggested that a 1 sd increase in the NPO corresponded with a 46% increase in GEP and ER at two western stations and decreases of 47% and 25% in GEP and ER, respectively, at east-central stations. Comparison of the scaled slope coefficients for GEP and ER from all relationships during spring indicated that the dynamical forcing associated with the NAO and NPO was about 55% stronger for GEP relative to ER. During summer, scaled slope coefficients indicated that the forcing was 35% stronger in GEP relative to ER.

Several studies have suggested that the NPO may be affected by other factors, such as ENSO, the AO/NAO, and anthropogenic warming (Folland et al. 2001). For this reason, it is important to address likely nonstationarities in the relationships reported here. Treating flux records from the north-central stations as independent samples, we regressed the standardized (z score) flux residuals (controlling for lag-1 autocorrelation) for all stations against SLP-2 during spring (Figs. 9a,b). The overall relationship in this region was significant for GEP and ER. By regressing these residuals onto SLP-1 (Figs. 9c,d), we show that the NAO explained a considerable amount of the unmodeled variability in the relationship between the NPO and spring fluxes. Figures 9e and 9f show the same residual analysis but with standardized P. This example offers insight into the potential sources of nonstationary behavior in the relationships between individual modes of large-scale tropospheric circulation variability and surface variables, such as GEP and ER. In this case, it is clear that the NAO acted to modulate the strength of the NPO’s effect over spring fluxes in the north-central region. Thus, low-frequency fluctuations in the state of the NAO may introduce similar low-frequency fluctuations in the magnitude of the NPO–flux relationship. We expected that a parallel analysis of the partial correlations between flux residuals and P during summer would show a similar modulation of relationships by explaining local variability of P. Although P generally failed to explain significant variation in summer flux residuals derived from the partial correlation analysis of the relationships with the NAO or NPO, we are hesitant to dismiss the role of local hydrology in causing periodic outliers in mode–flux relationships. Future analyses would benefit greatly by developing hydrological time series that more accurately reflect local moisture availability. In light of these results, it is suggested that large-scale modes of variability, such as the AO/NAO and NPO, may act as principal controls on climate-induced variability in carbon cycling throughout large regions, while frequent perturbations in local hydrological conditions act as subsidiary, but nontrivial, deviations from the large-scale relationships.

In many cases, fluxes were also correlated with modes during the preceding season (Tables 6 and 7). In nearly all cases, these lag effects were also associated with lag correlation in SAT and/or P, indicating that the relationships were caused by temporal dependence in atmospheric circulation, as opposed to complex land surface feedbacks. Composite maps, similar to those shown in Figs. 5 and 6, but illustrating the lag-1 response following high and low values from the preceding season, showed that patterns resembling the NAO and NPO did indeed persist from winter to spring and from spring to summer, despite an absence of lag-1 autocorrelation in the seasonal time series of expansion coefficients.

In summary, variability in seasonal GEP and ER has been linked to the immediate and preceding state of large-scale tropospheric circulation over North America. Correlation across the station network was spatially coherent and generally consistent with large-scale synoptic conditions. Some differences between neighboring stations could be attributed to processes not captured by large-scale climate, such as serial dependence, local climate, plant functional type, and hydrological conditions. While this suggests that large-scale modes of variability may influence the carbon cycle of these regions, improved (site-specific) knowledge of nonsystematic errors and longer EC time series are required to assess the robustness of the relationships over longer time scales. Thus, long-term monitoring EC stations show promise in characterizing the regional and ecosystem-specific carbon cycle response to low-frequency modes of tropospheric circulation variability and may play a critical role in validating ecosystem model responses to such phenomena.

Acknowledgments

This research was supported by funds awarded to P. Lafleur under the Fluxnet Canada Research Network from National Sciences and Engineering Research Council of Canada, Canadian Foundation for Climate and Atmospheric Sciences, and the Biocap Canada Foundation. With the exception of NTTP station, flux data were downloaded from the Fluxnet Canada and the Ameriflux Web sites. We acknowledge the following principal investigators for their hard work and dedication to collection and archiving of these data: A. Black (WTDF and SBAH), A. Barr (SBAH and SBBS), L. Flanagan (NTTP), D. Hollinger (EBSH), H. McCaughey (SBJP), S. Wofsy (NBBS and ETOM), G. Katul and R. Oren (ETOH and ETPP), R. Monson (WAFS), H. Schmid (STMT), and B. Cook and K. Davis (NTAM). We also recognize the hard work of student and associate members in the maintenance and operation of station measurements. NCEP–NCAR reanalysis and North American Regional Reanalysis data were provided by the NOAA–CIRES Climate Diagnostics Center, Boulder, Colorado. We also appreciate critical comments made by Dr. J. G. Cogley, Dr. L. Flanagan, and two anonymous reviewers.

REFERENCES

  • Adams, J. M., , and G. Piovesan, 2005: Long series relationships between global interannual CO2 increment and climate: Evidence for stability and change in role of the tropical and boreal-temperate zones. Chemosphere, 59 , 15951612.

    • Search Google Scholar
    • Export Citation
  • Baldocchi, D. D., , B. B. Hicks, , and T. P. Meyers, 1988: Measuring biosphere-atmosphere exchanges of biologically related gases with micrometeorological methods. Ecology, 69 , 13311340.

    • Search Google Scholar
    • Export Citation
  • Barford, C. C., and Coauthors, 2001: Factors controlling long- and short-term sequestration of atmospheric CO2 in a mid-latitude forest. Science, 294 , 16881691.

    • Search Google Scholar
    • Export Citation
  • Barnston, A. G., , and R. E. Livezey, 1987: Classification, seasonality, and persistence of low-frequency atmospheric circulation patterns. Mon. Wea. Rev., 115 , 10831126.

    • Search Google Scholar
    • Export Citation
  • Barr, A. G., , T. J. Griffis, , T. A. Black, , X. Lee, , R. M. Staebler, , J. D. Fuentes, , Z. Chen, , and K. Morgenstern, 2002: Comparing the carbon budgets of boreal and temperate deciduous forest stands. Can. J. For. Res., 32 , 813822.

    • Search Google Scholar
    • Export Citation
  • Black, T. A., and Coauthors, 2000: Increased carbon sequestration by a boreal deciduous forest in years with a warm spring. Geophys. Res. Lett., 27 , 12711274.

    • Search Google Scholar
    • Export Citation
  • Bousquet, P., , P. Peylin, , P. Ciais, , C. Le Quéré, , P. Friedlingtein, , and P. P. Tans, 2000: Regional changes in carbon dioxide fluxes of land and oceans since 1980. Science, 290 , 13421346.

    • Search Google Scholar
    • Export Citation
  • Braswell, B. H., , D. S. Schimel, , E. Linder, , and B. Moore III, 1997: The response of global terrestrial ecosystems to interannual temperature variability. Science, 278 , 870872.

    • Search Google Scholar
    • Export Citation
  • Buermann, W., , B. Anderson, , C. J. Tucker, , R. E. Dickinson, , W. Lucht, , C. S. Potter, , and R. B. Myneni, 2003: Interannual covariability in Northern Hemisphere air temperatures and greenness associated with El Niño-Southern Oscillation and the Arctic Oscillation. J. Geophys. Res., 108 .4396, doi:10.1029/2002JD002630.

    • Search Google Scholar
    • Export Citation
  • Cao, M., , and F. I. Woodward, 1998: Dynamic responses of terrestrial ecosystem carbon cycling to global climate change. Nature, 393 , 249252.

    • Search Google Scholar
    • Export Citation
  • Ciais, P., and Coauthors, 2005: Europe-wide reduction in primary productivity caused by the heat and drought in 2003. Nature, 437 , 529533.

    • Search Google Scholar
    • Export Citation
  • Cook, B. D., and Coauthors, 2004: Carbon exchange and venting anomalies in an upland deciduous forest in northern Wisconsin, USA. Agric. For. Meteor., 126 , 271295.

    • Search Google Scholar
    • Export Citation
  • Cramer, W., and Coauthors, 2001: Global response of terrestrial ecosystem structure and function to CO2 and climate change: Results from six dynamic global vegetation models. Global Change Biol., 7 , 357373.

    • Search Google Scholar
    • Export Citation
  • Dargaville, R. J., , S. C. Doney, , and I. Y. Fung, 2003: Inter-annual variability in the interhemispheric atmospheric CO2 gradient: Contributions from transport and the seasonal rectifier. Tellus, 55B , 711722.

    • Search Google Scholar
    • Export Citation
  • Deser, C., , and S. Phillips, 2006: Simulation of the 1976/77 climate transition over the North Pacific: Sensitivity to tropical forcing. J. Climate, 19 , 61706179.

    • Search Google Scholar
    • Export Citation
  • Dominguez, F., , and P. Kumar, 2005: Dominant modes of moisture flux anomalies over North America. J. Hydrometeor., 6 , 194209.

  • Dufresne, J-L., , P. Friedlingstein, , M. Berthelot, , L. Bopp, , P. Ciais, , L. Fairhead, , H. Le Treut, , and P. Monfray, 2002: On the magnitude of positive feedback between future climate change and the carbon cycle. Geophys. Res. Lett., 29 .1405, doi:10.1029/2001GL013777.

    • Search Google Scholar
    • Export Citation
  • Falge, E., and Coauthors, 2001: Gap filling strategies for defensible annual sums of net ecosystem exchange. Agric. For. Meteor., 107 , 4369.

    • Search Google Scholar
    • Export Citation
  • Favre, A., , and A. Gershunov, 2006: Extra-tropical cyclonic/anticyclonic activity in North-Eastern Pacific and air temperature extremes in Western North America. Climate Dyn., 26 , 617629.

    • Search Google Scholar
    • Export Citation
  • Flanagan, L. B., , L. A. Wever, , and P. J. Carlson, 2002: Seasonal and interannual variation in carbon dioxide exchange and carbon balance in a northern temperate grassland. Global Change Biol., 8 , 599615.

    • Search Google Scholar
    • Export Citation
  • Folland, C. K., and Coauthors, 2001: Observed climate variability and change. Climate Change 2001: The Scientific Basis, J. T. Houghton et al., Eds., Cambridge University Press, 99–182.

    • Search Google Scholar
    • Export Citation
  • Friedlingstein, P., and Coauthors, 2006: Climate–carbon cycle feedback analysis: Results from the C4MIP model intercomparison. J. Climate, 19 , 33373353.

    • Search Google Scholar
    • Export Citation
  • Fung, I. Y., , S. C. Doney, , K. Lindsay, , and J. John, 2005: Evolution of carbon sinks in a changing climate. Proc. Natl. Acad. Sci. USA, 102 , 1120111206.

    • Search Google Scholar
    • Export Citation
  • Goulden, M. L., , J. W. Munger, , S-M. Fan, , B. C. Daube, , and S. C. Wofsy, 1996: Measurements of carbon sequestration by long-term eddy covariance: Methods and a critical evaluation of accuracy. Global Change Biol., 2 , 169182.

    • Search Google Scholar
    • Export Citation
  • Goulden, M. L., , B. C. Daube, , S-M. Fan, , D. J. Sutton, , A. Bazzaz, , J. W. Munger, , and S. C. Wofsy, 1997: Physiological responses of a black spruce forest to weather. J. Geophys. Res., 102 , D24. 2898728996.

    • Search Google Scholar
    • Export Citation
  • Griffis, T. J., , T. A. Black, , K. Morgenstern, , A. G. Barr, , Z. Nesic, , G. B. Drewitt, , D. Gaumont-Guay, , and J. H. McCaughey, 2003: Ecophysiological controls on the carbon balances of three southern boreal forests. Agric. For. Meteor., 117 , 5371.

    • Search Google Scholar
    • Export Citation
  • Gu, L., and Coauthors, 2005: Objective threshold determination for nighttime eddy flux filtering. Agric. For. Meteor., 128 , 179197.

  • Gurney, K. R., and Coauthors, 2002: Towards robust regional estimates of CO2 sources and sinks using atmospheric transport models. Nature, 415 , 626630.

    • Search Google Scholar
    • Export Citation
  • Hember, R. A., , P. M. Lafleur, , and J. G. Cogley, 2005: Synoptic controls on summer evapotranspiration from a bog peatland in southern Canada. Int. J. Climatol., 25 , 793809.

    • Search Google Scholar
    • Export Citation
  • Hollinger, D. Y., , and A. D. Richardson, 2005: Uncertainty in eddy covariance measurements and its application to physiological models. Tree Physiol., 25 , 873885.

    • Search Google Scholar
    • Export Citation
  • Hollinger, D. Y., , S. M. Goltz, , E. A. Davidson, , J. T. Lee, , K. Tu, , and H. T. Valentine, 1999: Seasonal patterns and environmental controls of carbon dioxide and water vapour exchange in an ecotonal boreal forest. Global Change Biol., 5 , 891902.

    • Search Google Scholar
    • Export Citation
  • Hollinger, D. Y., and Coauthors, 2004: Spatial and temporal variability in forest–atmosphere CO2 exchange. Global Change Biol., 10 , 16891706.

    • Search Google Scholar
    • Export Citation
  • Houghton, R. A., 2003: Why are estimates of the terrestrial carbon balance so different? Global Change Biol., 9 , 500509.

  • House, J. I., , I. C. Prentice, , N. Ramankutty, , R. A. Houghton, , and M. Heimann, 2003: Reconciling apparent inconsistencies in estimates of terrestrial CO2 sources and sinks. Tellus, 55B , 345363.

    • Search Google Scholar
    • Export Citation
  • Humphreys, E. R., , T. A. Black, , K. Morgenstern, , Z. Li, , and Z. Nesic, 2005: Net ecosystem production of a Douglas-fir stand for 3 years following clearcut harvesting. Global Change Biol., 11 , 450464.

    • Search Google Scholar
    • Export Citation
  • Hurrell, D. A., , and H. van Loon, 1997: Decadal variations in climate associated with the North Atlantic Oscillation. Climatic Change, 36 , 301326.

    • Search Google Scholar
    • Export Citation
  • Janssens, I. A., and Coauthors, 2001: Productivity overshadows temperature in determining soil and ecosystem respiration across European forests. Global Change Biol., 7 , 269278.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77 , 437471.

  • Lafleur, P. M., , N. T. Roulet, , and S. W. Admiral, 2001: Annual cycle of CO2 exchange at a bog peatland. J. Geophys. Res., 106 , D3. 30713081.

    • Search Google Scholar
    • Export Citation
  • Lafleur, P. M., , N. T. Roulet, , J. L. Bubier, , S. Frolking, , and T. R. Moore, 2003: Interannual variability in the peatland-atmosphere carbon dioxide exchange at an ombrotrophic bog. Global Biogeochem. Cycles, 17 .1036, doi:10.1029/2002GB001983.

    • Search Google Scholar
    • Export Citation
  • Limpasuvan, V., , and D. L. Hartmann, 1999: Eddies and the annular modes of climate variability. Geophys. Res. Lett., 26 , 31333136.

  • Liu, J., , R. E. Stewart, , and K. K. Szeto, 2004: Moisture transport and other hydrometeorological features associated with the severe 2000/01 drought over the western and central Canadian Prairies. J. Climate, 17 , 305319.

    • Search Google Scholar
    • Export Citation
  • Massman, W. J., , and X. Lee, 2002: Eddy covariance flux corrections and uncertainties in long-term studies of carbon and energy exchanges. Agric. For. Meteor., 113 , 121144.

    • Search Google Scholar
    • Export Citation
  • Moncrieff, J. B., , Y. Malhi, , and R. Leuning, 1996: The propagation of errors in long-term measurements of land-atmosphere fluxes of carbon and water. Global Change Biol., 2 , 231240.

    • Search Google Scholar
    • Export Citation
  • Morgenstern, K., and Coauthors, 2004: Sensitivity and uncertainty of the carbon balance of a Pacific Northwest Douglas-fir forest during an El Niño/La Niña cycle. Agric. For. Meteor., 123 , 201219.

    • Search Google Scholar
    • Export Citation
  • Newman, M., , G. P. Compo, , and M. A. Alexander, 2003: ENSO-forced variability of the Pacific decadal oscillation. J. Climate, 16 , 38533857.

    • Search Google Scholar
    • Export Citation
  • North, G. R., , T. L. Bell, , R. F. Cahalan, , and F. J. Moeng, 1982: Sampling errors in the estimation of empirical orthogonal functions. Mon. Wea. Rev., 110 , 699706.

    • Search Google Scholar
    • Export Citation
  • Ogi, M., , K. Yamazaki, , and Y. Tachibana, 2004: The summertime annular mode in the Northern Hemisphere and its linkage to the winter mode. J. Geophys. Res., 109 .D20114, doi:10.1029/2004JD004514.

    • Search Google Scholar
    • Export Citation
  • Ogi, M., , K. Yamazaki, , and Y. Tachibana, 2005: The summer northern annular mode and abnormal summer weather in 2003. Geophys. Res. Lett., 32 .L04706, doi:10.1029/2004GL021528.

    • Search Google Scholar
    • Export Citation
  • Palmroth, S., , C. A. Maier, , H. R. McCarthy, , A. C. Oishi, , H-S. Kim, , K. H. Johnsen, , G. G. Katul, , and R. Oren, 2005: Contrasting responses to drought of forest floor CO2 efflux in a Loblolly pine plantation and a nearby Oak-Hickory forest. Global Change Biol., 11 , 421434.

    • Search Google Scholar
    • Export Citation
  • Patra, P. K., , M. Ishizawa, , S. Maksyutov, , T. Nakazawa, , and G. Inoue, 2005: Role of biomass burning and climate anomalies for land-atmosphere carbon fluxes based on inverse modeling of atmospheric CO2. Global Biogeochem. Cycles, 19 .GB3005, doi:10.1029/2004GB002258.

    • Search Google Scholar
    • Export Citation
  • Peylin, P., and Coauthors, 2005: Multiple constraints on regional CO2 flux variations over land and ocean. Global Biogeochem. Cycles, 19 .GB1011, doi:10.1029/2003GB002214.

    • Search Google Scholar
    • Export Citation
  • Ponton, S., , L. B. Flanagan, , K. P. Alstad, , B. G. Johnson, , K. Morgenstern, , N. Kljun, , T. A. Black, , and A. G. Barr, 2006: Comparison of ecosystem water-use efficiency among Douglas-fir forest, aspen forest, and grassland using eddy covariance and carbon isotope techniques. Global Change Biol., 12 , 294310.

    • Search Google Scholar
    • Export Citation
  • Press, W. H., , B. P. Flannery, , S. A. Teukolsky, , and W. T. Vetterling, 1986: Numerical Recipes: The Art of Scientific Computing. Cambridge University Press, 498 pp.

    • Search Google Scholar
    • Export Citation
  • Raich, J. W., , and W. H. Schlesinger, 1992: The global carbon dioxide flux in soil respiration and its relationship to vegetation and climate. Tellus, 44B , 8199.

    • Search Google Scholar
    • Export Citation
  • Richardson, A. D., , and D. Y. Hollinger, 2005: Statistical modeling of ecosystem respiration using eddy covariance data: Maximum likelihood parameter estimation, and Monte Carlo simulation of model and parameter uncertainty, applied to three simple models. Agric. For. Meteor., 131 , 191208.

    • Search Google Scholar
    • Export Citation
  • Richman, M. B., 1986: Rotation of principal components. Int. J. Climatol., 6 , 293335.

  • Rogers, J. C., 1997: North Atlantic storm track variability and its association to the North Atlantic Oscillation and climate variability of northern Europe. J. Climate, 10 , 16351647.

    • Search Google Scholar
    • Export Citation
  • Sacks, W. J., , D. S. Schimel, , R. K. Monson, , and B. H. Braswell, 2006: Model-data synthesis of diurnal and seasonal CO2 fluxes at Niwot Ridge, Colorado. Global Change Biol., 12 , 240259.

    • Search Google Scholar
    • Export Citation
  • Sakai, R. K., , D. R. Fitzjarrald, , and K. E. Moore, 2001: Importance of low-frequency contributions to eddy fluxes observed over rough surfaces. J. Appl. Meteor., 40 , 21782192.

    • Search Google Scholar
    • Export Citation
  • Schaefer, K., , A. S. Denning, , N. Suits, , J. Kaduk, , I. Baker, , S. Los, , and L. Prihodko, 2002: Effect of climate on interannual variability of terrestrial CO2 fluxes. Global Biogeochem. Cycles, 16 .1101, doi:10.1029/2002GB001928.

    • Search Google Scholar
    • Export Citation
  • Schmid, H. P., , C. S. Grimmond, , F. Cropley, , B. Offerle, , and H-B. Su, 2000: Measurements of CO2 and energy fluxes over a mixed hardwood forest in the midwestern United States. Agric. For. Meteor., 103 , 357374.

    • Search Google Scholar
    • Export Citation
  • Schmid, H. P., , H-B. Su, , C. S. Vogel, , and P. S. Curtis, 2003: Ecosystem-atmosphere exchange of carbon dioxide over a mixed hardwood forest in northern lower Michigan. J. Geophys. Res., 108 .4417, doi:10.1029/2002JD003011.

    • Search Google Scholar
    • Export Citation
  • Shindell, D. T., , G. A. Schmidt, , M. E. Mann, , D. Rind, , and A. Waple, 2001: Solar forcing of regional climate change during the Maunder Minimum. Science, 294 , 21492152.

    • Search Google Scholar
    • Export Citation
  • Stocker, T. F., and Coauthors, 2001: Physical climate processes and feedbacks. Climate Change 2001: The Scientific Basis, J. T. Houghton et al., Eds., Cambridge University Press, 417–470.

    • Search Google Scholar
    • Export Citation
  • Stoy, P. C., , G. G. Katul, , M. B. S. Siqueira, , J-Y. Juang, , K. A. Novick, , J. M. Uebelherr, , and R. Oren, 2006: An evaluation of models for partitioning eddy covariance-measured net ecosystem exchange into photosynthesis and respiration. Agric. For. Meteor., 141 , 218.

    • Search Google Scholar
    • Export Citation
  • Thompson, D. W. J., , and J. M. Wallace, 1998: The Arctic Oscillation signature in the wintertime geopotential height and temperature fields. Geophys. Res. Lett., 25 , 12971300.

    • Search Google Scholar
    • Export Citation
  • Thompson, D. W. J., , and J. M. Wallace, 2000: Annular modes in the extratropical circulation. Part I: Month-to-month variability. J. Climate, 13 , 10001016.

    • Search Google Scholar
    • Export Citation
  • Thompson, D. W. J., , and J. M. Wallace, 2001: Regional climate impacts of the Northern Hemisphere annular mode. Science, 293 , 8689.

  • Thompson, D. W. J., , J. M. Wallace, , and G. C. Hegerl, 2000: Annular modes in the extratropical circulation. Part II: Trends. J. Climate, 13 , 10181036.

    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., , G. W. Branstator, , D. Karoly, , A. Kumar, , N-C. Lau, , and C. Ropelewski, 1998: Progress during TOGA in understanding and modeling global teleconnections associated with tropical sea surface temperatures. J. Geophys. Res., 103 , C7. 1429114324.

    • Search Google Scholar
    • Export Citation
  • Twine, T. E., and Coauthors, 2000: Correcting eddy-covariance flux underestimates over a grassland. Agric. For. Meteor., 103 , 279300.

    • Search Google Scholar
    • Export Citation
  • von Storch, H., , and F. W. Zwiers, 1999: Statistical Analysis in Climate Research. Cambridge University Press, 484 pp.

  • Wallace, J. M., , and D. S. Gutzler, 1981: Teleconnections in the geopotential height field during the Northern Hemisphere winter. Mon. Wea. Rev., 109 , 784812.

    • Search Google Scholar
    • Export Citation
  • Wallace, J. M., , Y. Zhang, , and J. A. Renwick, 1995: Dynamic contribution to hemispheric mean temperature trends. Science, 270 , 780783.

    • Search Google Scholar
    • Export Citation
  • Wallace, J. M., , Y. Zhang, , and L. Bajuk, 1996: Interpretation of interdecadal trends in Northern Hemisphere surface air temperature. J. Climate, 9 , 249260.

    • Search Google Scholar
    • Export Citation
  • Wang, X. L., , H. Wan, , and V. R. Swail, 2006: Observed changes in cyclone activity in Canada and their relationships to major circulation regimes. J. Climate, 19 , 896915.

    • Search Google Scholar
    • Export Citation
  • Wilson, K., and Coauthors, 2002: Energy balance closure at FLUXNET sites. Agric. For. Meteor., 113 , 223243.

  • Wu, Q., , and D. M. Straus, 2004: AO, COWL, and observed climate trends. J. Climate, 17 , 21392156.

  • Zeng, N., , A. Mariotti, , and P. Wetzel, 2005a: Terrestrial mechanisms of interannual CO2 variability. Global Biogeochem. Cycles, 19 .GB1016, doi:10.1029/2004GB002273.

    • Search Google Scholar
    • Export Citation
  • Zeng, N., , H. Qian, , C. Roedenbeck, , and M. Heimann, 2005b: Impact of 1998–2002 midlatitude drought and warming on terrestrial ecosystem and the global carbon cycle. Geophys. Res. Lett., 32 .L22709, doi:10.1029/2005GL024607.

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

Locations of eddy covariance stations and time series of cumulative monthly NEP since January 1994. Monthly averages have been used as placeholders for periods of missing data.

Citation: Journal of Climate 21, 4; 10.1175/2007JCLI1718.1

Fig. 2.
Fig. 2.

Experimental surface of uncertainty in measurements of ER [σ(δ)] based on the standard deviation of first-order differences in hourly nighttime elements, binned according to friction velocity u* and soil temperature Tsoil at the Mer Bleue bog (station SBBP).

Citation: Journal of Climate 21, 4; 10.1175/2007JCLI1718.1

Fig. 3.
Fig. 3.

First rotated empirical orthogonal function for the (a) SLP pattern, (b) Z500 pattern, (c) standardized seasonal time series, and (d) monthly standard deviation for expansion coefficients. The boldface solid contour equals zero, thin solid contours indicate positive anomalies, and thin, broken contours indicate negative anomalies. SLP contour interval = 0.5 hPa. Z500 contour interval = 20 m.

Citation: Journal of Climate 21, 4; 10.1175/2007JCLI1718.1

Fig. 4.
Fig. 4.

As in Fig. 3 but for the second rotated EOF.

Citation: Journal of Climate 21, 4; 10.1175/2007JCLI1718.1

Fig. 5.
Fig. 5.

Composite maps illustrating the phases of (a), (b) SLP-1 during MAM; (c), (d) Z500-1 during MAM; (e), (f) SLP-1 during JJA; and (g), (h) Z500-1 during JJA. Filled contours indicate SAT composite anomalies (0.5°C intervals). Superimposed contours indicate composite anomalies of SLP (0.5-hPa intervals) or Z500 (20-m intervals).

Citation: Journal of Climate 21, 4; 10.1175/2007JCLI1718.1

Fig. 6.
Fig. 6.

Composite maps illustrating the phases of (a), (b) SLP-2 during MAM; (c), (d) Z500-2 during MAM; (e), (f) SLP-2 during JJA; and (g), (h) Z500-2 during JJA. Filled contours indicate SAT composite anomalies (0.5°C intervals). Superimposed contours indicate composite anomalies of SLP (0.5-hPa intervals) or Z500 (20-m intervals).

Citation: Journal of Climate 21, 4; 10.1175/2007JCLI1718.1

Fig. 7.
Fig. 7.

Correlation coefficients (×100) from regressing station variables onto expansion coefficients of (a) spring SLP-1, (b) spring Z500-1, (c) summer SLP-1, and (d) summer Z500-1. Upper-left box: SAT; upper-right box: P; lower-left box: GEP; lower-right box: ER. Boldface coefficients indicate significance at the 95% confidence level.

Citation: Journal of Climate 21, 4; 10.1175/2007JCLI1718.1

Fig. 8.
Fig. 8.

As in Fig. 7, but for (a) spring SLP-2, (b) spring Z500-2, (c) summer SLP-2, and (d) summer Z500-2.

Citation: Journal of Climate 21, 4; 10.1175/2007JCLI1718.1

Fig. 9.
Fig. 9.

Residual analysis of the (a), (b) relationships between spring SLP-2 and standardized CO2 flux residuals after controlling lag-1 autocorrelation; the (c), (d) relationships between spring SLP-1 and residuals from (a) and (b), and the (e), (f) relationships between spring P and residuals from (a) and (b).

Citation: Journal of Climate 21, 4; 10.1175/2007JCLI1718.1

Table 1.

Eddy covariance station descriptions. Elevations given in meters above sea level (m ASL).

Table 1.
Table 2.

Comparison of monthly P from EC station gauge measurements and the nearest grid cell value of the NARR dataset. Here, N indicates the number of measurements (sample size).

Table 2.
Table 3.

Evaluation of lag-1 autocorrelation and long-term trends in CO2 fluxes: slope coefficients (b1) are in g C m−2. Boldface r2 values mark statistical significance at the 95% confidence level. Spring GEP was regressed onto winter ER as opposed to winter GEP.

Table 3.
Table 4.

Partial cross correlation between CO2 fluxes and the first EOFs, while controlling for lag-1 serial autocorrelation in CO2 fluxes. Spring GEP residuals were derived from regression with winter ER. Boldface r2 values mark significance at the 95% confidence level. Slope coefficients (b1) indicate the CO2 flux residual response to a unit change (+1 sd) in the forcing variable (g C m−2).

Table 4.
Table 5.

Partial cross correlation between CO2 fluxes and the second EOFs, while controlling for lag-1 serial autocorrelation in CO2 fluxes. Spring GEP residuals were derived from regression with winter ER. Boldface r2 values mark significance at the 95% confidence level. Slope coefficients (b1) indicate the CO2 flux residual response to a unit change (+1 sd) in the forcing variable (g C m−2).

Table 5.
Table 6.

Slope coefficients from lag-1 linear regression analysis of station variables and the first EOF during spring and summer. Coefficients express the response to a unit change (±1 sd) in the NAO: SAT (°C), P (mm), GEP (g C m−2), and ER (g C m−2). Boldface values indicate significance at the 95% confidence level.

Table 6.
Table 7.

Slope coefficients from lag-1 linear regression analysis of station variables and the second EOF during spring and summer. Coefficients express the response to a unit change (±1 sd) in the NPO: SAT (°C), P (mm), GEP (g C m−2), and ER (g C m−2). Boldface values indicate significance at the 95% confidence level.

Table 7.
Save