Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

A two-fold increase of carbon cycle sensitivity to tropical temperature variations


This article has been updated


Earth system models project that the tropical land carbon sink will decrease in size in response to an increase in warming and drought during this century, probably causing a positive climate feedback1,2. But available data3,4,5 are too limited at present to test the predicted changes in the tropical carbon balance in response to climate change. Long-term atmospheric carbon dioxide data provide a global record that integrates the interannual variability of the global carbon balance. Multiple lines of evidence6,7,8 demonstrate that most of this variability originates in the terrestrial biosphere. In particular, the year-to-year variations in the atmospheric carbon dioxide growth rate (CGR) are thought to be the result of fluctuations in the carbon fluxes of tropical land areas6,9,10. Recently, the response of CGR to tropical climate interannual variability was used to put a constraint on the sensitivity of tropical land carbon to climate change10. Here we use the long-term CGR record from Mauna Loa and the South Pole to show that the sensitivity of CGR to tropical temperature interannual variability has increased by a factor of 1.9 ± 0.3 in the past five decades. We find that this sensitivity was greater when tropical land regions experienced drier conditions. This suggests that the sensitivity of CGR to interannual temperature variations is regulated by moisture conditions, even though the direct correlation between CGR and tropical precipitation is weak9. We also find that present terrestrial carbon cycle models do not capture the observed enhancement in CGR sensitivity in the past five decades. More realistic model predictions of future carbon cycle and climate feedbacks require a better understanding of the processes driving the response of tropical ecosystems to drought and warming.

This is a preview of subscription content, access via your institution

Relevant articles

Open Access articles citing this article.

Access options

Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Figure 1: Change in detrended anomalies in CGR and tropical MAT, in dCGR/dMAT and in over the past five decades.
Figure 2: Histograms of during the earliest period and during the most recent period, and of relative change in over the past five decades.
Figure 3: for each bin of detrended tropical moisture anomaly.

Change history

  • 12 February 2014

    Reference 31 has been moved to the main-text reference list.


  1. Cox, P. M., Betts, R. A., Jones, C. D., Spall, S. A. & Totterdell, I. J. Acceleration of global warming due to carbon-cycle feedbacks in a coupled climate model. Nature 408, 184–187 (2000)

    ADS  CAS  PubMed  Google Scholar 

  2. Sitch, S. et al. Evaluation of the terrestrial carbon cycle, future plant geography and climate-carbon cycle feedbacks using five dynamic global vegetation models (DGVMs). Glob. Change Biol. 14, 2015–2039 (2008)

    ADS  Google Scholar 

  3. Phillips, O. L. et al. Drought sensitivity of the Amazon rainforest. Science 323, 1344–1347 (2009)

    ADS  CAS  PubMed  Google Scholar 

  4. Meir, P. & Woodward, F. I. Amazonian rain forests and drought: response and vulnerability. New Phytol. 187, 553–557 (2010)

    PubMed  Google Scholar 

  5. Davidson, E. A. et al. The Amazon basin in transition. Nature 481, 321–328 (2012)

    ADS  CAS  PubMed  Google Scholar 

  6. Baker, D. F. et al. TransCom 3 inversion intercomparison: Impact of transport model errors on the interannual variability of regional CO2 fluxes, 1988–2003. Glob. Biogeochem. Cycles 20, GB1002 (2006)

    ADS  Google Scholar 

  7. Lee, K., Wanninkhof, R., Takahashi, T., Doney, S. C. & Feely, R. A. Low interannual variability in recent oceanic uptake of atmospheric carbon dioxide. Nature 396, 155–159 (1998)

    ADS  CAS  Google Scholar 

  8. Alden, C. B., Miller, J. B. & White, J. W. C. Can bottom-up ocean CO2 fluxes be reconciled with atmospheric 13C observations? Tellus B 62, 369–388 (2010)

    ADS  Google Scholar 

  9. Wang, W. et al. Variations in atmospheric CO2 growth rates coupled with tropical temperature. Proc. Natl Acad. Sci. USA 110, 13061–13066 (2013)

    ADS  CAS  PubMed  Google Scholar 

  10. Cox, P. M. et al. Sensitivity of tropical carbon to climate change constrained by carbon dioxide variability. Nature 494, 341–344 (2013)

    ADS  CAS  PubMed  Google Scholar 

  11. Patra, P. K., Ishizawa, M., Maksyutov, S., Nakazawa, T. & Inoue, G. Role of biomass burning and climate anomalies for land-atmosphere carbon fluxes based on inverse modeling of atmospheric CO2 . Glob. Biogeochem. Cycles 19, GB3005 (2005)

    ADS  Google Scholar 

  12. Keeling, C. D., Whorf, T. P., Wahlen, M. & van der Plichtt, J. Interannual extremes in the rate of rise of atmospheric carbon dioxide since 1980. Nature 375, 666–670 (1995)

    ADS  CAS  Google Scholar 

  13. Adams, J. M. & Piovesan, G. 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, 1595–1612 (2005)

    ADS  CAS  PubMed  Google Scholar 

  14. Corlett, R. T. Impacts of warming on tropical lowland rainforests. Trends Ecol. Evol. 26, 606–613 (2011)

    PubMed  Google Scholar 

  15. Sugihara, G. et al. Detecting causality in complex ecosystems. Science 338, 496–500 (2012)

    ADS  CAS  PubMed  MATH  Google Scholar 

  16. Mitchell, T. D. & Jones, P. D. An improved method of constructing a database of monthly climate observations and associated high-resolution grids. Int. J. Climatol. 25, 693–712 (2005)

    Google Scholar 

  17. Sarmiento, J. L. et al. Trends and regional distributions of land and ocean carbon sinks. Biogeosciences 7, 2351–2367 (2010)

    ADS  CAS  Google Scholar 

  18. Mahecha, M. D. et al. Characterizing ecosystem-atmosphere interactions from short to interannual time scales. Biogeosciences 4, 743–758 (2007)

    ADS  CAS  Google Scholar 

  19. Smith, T. M., Reynolds, R. W., Peterson, T. C. & Lawrimore, J. Improvements to NOAA’s historical merged land–ocean surface temperature analysis (1880–2006). J. Clim. 21, 2283–2296 (2008)

    ADS  Google Scholar 

  20. Schneider, U. et al. GPCC’s new land surface precipitation climatology based on quality-controlled in situ data and its role in quantifying the global water cycle. Theor. Appl. Climatol.. (2013)

  21. Sheffield, J., Goteti, G. & Wood, E. F. Development of a 50-year high-resolution global dataset of meteorological forcings for land surface modeling. J. Clim. 19, 3088–3111 (2006)

    ADS  Google Scholar 

  22. Le Quéré, C. et al. Trends in the sources and sinks of carbon dioxide. Nature Geosci. 2, 831–836 (2009)

    ADS  Google Scholar 

  23. Dai, A. Characteristics and trends in various forms of the Palmer drought severity index during 1900–2008. J. Geophys. Res. 116, D12115 (2011)

    ADS  Google Scholar 

  24. Vicente-Serrano, S. M., Beguería, S. & López-Moreno, J. I. A multiscalar drought index sensitive to global warming: the standardized precipitation evapotranspiration index. J. Clim. 23, 1696–1718 (2010)

    ADS  Google Scholar 

  25. Piao, S. et al. Spatiotemporal patterns of terrestrial carbon cycle during the 20th century. Glob. Biogeochem. Cycles 23, GB4026 (2009)

    ADS  Google Scholar 

  26. Schwalm, C. R. et al. Does terrestrial drought explain global CO2 flux anomalies induced by El Niño? Biogeosciences 8, 2493–2506 (2011)

    ADS  CAS  Google Scholar 

  27. Dai, A. Increasing drought under global warming in observations and models. Nature Clim. Change 3, 52–58 (2013)

    ADS  Google Scholar 

  28. Randerson, J. T. Climate science: global warming and tropical carbon. Nature 494, 319–320 (2013)

    ADS  CAS  PubMed  Google Scholar 

  29. Ballantyne, A. P., Alden, C. B., Miller, J. B., Tans, P. P. & White, J. W. C. Increase in observed net carbon dioxide uptake by land and oceans during the past 50 years. Nature 488, 70–72 (2012)

    ADS  CAS  PubMed  Google Scholar 

  30. Graven, H. D. et al. Enhanced seasonal exchange of CO2 by northern ecosystems since 1960. Science 341, 1085–1089 (2013)

    ADS  CAS  PubMed  Google Scholar 

  31. Keeling, C. D. et al. in A History of Atmospheric CO 2 and its effects on Plants, Animals, and Ecosystems (eds Ehleringer, J. R. et al.) 83–113 (Springer, 2005)

  32. Kistler, R. et al. The NCEP–NCAR 50-year reanalysis: monthly means CD–ROM and documentation. Bull. Am. Meteorol. Soc. 82, 247–267 (2001)

    ADS  Google Scholar 

  33. Angstrom, A. Solar and terrestrial radiation: report to the international commission for solar research on actinometric investigations of solar and atmospheric radiation. Q. J. R. Meteorol. Soc. 50, 121–126 (1924)

    ADS  Google Scholar 

  34. Stackhouse, P. et al. 12-year surface radiation budget data set. GEWEX News 14, 10–12 (2004)

    Google Scholar 

  35. Dinku, T., Connor, S. J., Ceccato, P. & Ropelewski, C. F. Comparison of global gridded precipitation products over a mountainous region of Africa. Int. J. Climatol. 28, 1627–1638 (2008)

    Google Scholar 

  36. Solomon S., et al., eds. (eds) Climate Change 2007: The Physical Science Basis 241–265 (Cambridge Univ. Press, 2007)

  37. Sheffield, J., Wood, E. F. & Roderick, M. L. Little change in global drought over the past 60 years. Nature 491, 435–438 (2012)

    ADS  CAS  PubMed  Google Scholar 

  38. Seneviratne, S. I. et al. Investigating soil moisture-climate interactions in a changing climate: a review. Earth Sci. Rev. 99, 125–161 (2010)

    ADS  CAS  Google Scholar 

  39. Edwards, D. C. Characteristics of 20th Century Drought in the United States at Multiple Time Scales. MSc thesis, Colorado State Univ. (1997)

  40. Hirschi, M. et al. Observational evidence for soil-moisture impact on hot extremes in southeastern Europe. Nature Geosci. 4, 17–21 (2011)

    ADS  CAS  Google Scholar 

  41. Vicente-Serrano, S. M., Beguería, S., López-Moreno, J. I., Angulo, M. & El Kenawy, A. A new global 0.5° gridded dataset (1901–2006) of a multiscalar drought index: comparison with current drought index datasets based on the Palmer drought severity index. J. Hydrometeorol. 11, 1033–1043 (2010)

    ADS  Google Scholar 

  42. Tucker, C. J. et al. An extended AVHRR 8-km NDVI dataset compatible with MODIS and SPOT vegetation NDVI data. Int. J. Remote Sens. 26, 4485–4498 (2005)

    ADS  Google Scholar 

  43. Clark, D. A., Piper, S. C., Keeling, C. D. & Clark, D. B. Tropical rain forest tree growth and atmospheric carbon dynamics linked to interannual temperature variation during 1984–2000. Proc. Natl Acad. Sci. USA 100, 5852–5857 (2003)

    ADS  CAS  PubMed  Google Scholar 

  44. van Mantgem, P. J. & Stephenson, N. L. Apparent climatically induced increase of tree mortality rates in a temperate forest. Ecol. Lett. 10, 909–916 (2007)

    PubMed  Google Scholar 

  45. Carnicer, J. et al. Widespread crown condition decline, food web disruption, and amplified tree mortality with increased climate change-type drought. Proc. Natl Acad. Sci. USA 108, 1474–1478 (2011)

    ADS  CAS  PubMed  Google Scholar 

  46. Mann, H. B. Nonparametric tests against trend. Econometrica 13, 245–259 (1945)

    MathSciNet  MATH  Google Scholar 

  47. Mahecha, M. D., Fürst, L. M., Gobron, N. & Lange, H. Identifying multiple spatiotemporal patterns: a refined view on terrestrial photosynthetic activity. Pattern Recognit. Lett. 31, 2309–2317 (2010)

    Google Scholar 

  48. Manly, B. Randomization, Bootstrap and Monte Carlo Methods in Biology 3rd edn (Chapman & Hall/CRC, 2007)

    MATH  Google Scholar 

  49. Solomon S., et al., eds. (eds) Climate Change 2001: The Scientific Basis 351–358 (Cambridge Univ. Press, 2007)

  50. Hansen, J. et al. Global climate changes as forecast by Goddard Institute for Space Studies three-dimensional model. J. Geophys. Res. 93, 9341–9364 (1988)

    ADS  CAS  Google Scholar 

  51. Alexander, L. V. et al. in Climate Change 2013: The Physical Science Basis (eds Stocker, T. F. et al.) 1–33 (Cambridge Univ. Press, 2013)

Download references


We thank S. Seneviratne and G. Zhang for comments and J. Gash for English editing. This study was supported by the National Natural Science Foundation of China (41125004), the National Basic Research Program of China (grant nos 2010CB950601 and 2013CB956303) and the National Youth Top-notch Talent Support Program in China. R.B.M. was funded by the NASA Earth Science Division.

Author information

Authors and Affiliations



S. Piao, P. Ciais and X.W. designed the research; X.W. performed the analysis; S. Piao, P. Ciais and X.W. drafted the paper; and P.F., R.B.M., P. Cox., M.H., J.M., S. Peng, T.W., H.Y. and A.C. contributed to the interpretation of the results and to the text.

Corresponding author

Correspondence to Shilong Piao.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Extended data figures and tables

Extended Data Figure 1 Spatial distribution of the correlation coefficient between detrended CGR and MAT anomalies.

CGR anomalies are from Mauna Loa Observatory and local MAT anomalies were derived from the CRU data set for the period 1960–2011. The correlation coefficients 0.23 and 0.28 are the critical thresholds at significance levels of 0.10 and 0.05 (n = 52), respectively.

Extended Data Figure 2 Convergent cross-mapping for reconstruction of variations in MAT, annual precipitation and mean annual solar radiation from variations in CGR.

CGR data are from Mauna Loa Observatory. The CGR-reconstructed temperature curve gradually converges to a large positive correlation coefficient (R = 0.70), whereas the CGR-reconstructed precipitation (P) and radiation (R) curves lead to smaller correlation coefficients (R = 0.04 and R = 0.23, respectively) as time-series length increases, suggesting that CGR variations are mainly forced by temperature variations rather than by variations in precipitation and solar radiation.

Extended Data Figure 3 Change in dCGR/dMAT, , and the effects of interannual variations of precipitation and solar radiation on the estimate of dCGR/dMAT.

The changes are calculated between the latest two decades and the earliest two decades in 1960–2011. Precipitation and radiation effects are denoted f2 × dP/dMAT and f3 × dR/dMAT, respectively. dCGR/dMAT is calculated as the slope of MAT in the regression of CGR at Mauna Loa Observatory against MAT over the tropical vegetated land. , f2 and f3 are the slopes of MAT, precipitation and radiation, respectively, in the multiple regression of CGR against MAT, precipitation and radiation over the tropical vegetated land. dP/dMAT is the slope of MAT in the regression of precipitation against MAT. dR/dMAT is the slope of MAT in the regression of radiation against MAT. Error bars indicate the 95% confidence interval of the corresponding value derived from 500 bootstrap estimates.

Extended Data Figure 4 and for the first and the last time window during the study period 1960–2011.

For 20-yr windows, the two windows are 1960–1979 and 1992–2011. For 25-yr windows, they are 1960–1984 and 1987–2011. a, Data from 1992 and 1993 (post-Pinatubo years) are excluded from estimates. b, Data from the record-high El Niño events of 1972–1973 and 1997–1998 are excluded from estimates. c, Interannual variations of CGR and climate variables are obtained from the frequency decomposition by SSA (Methods). d, is estimated with alternative climate data sets (tropical MAT is from the GHCN data set19, tropical annual precipitation is from the GPCC20 and solar radiation is from ref. 21). e, CGR is obtained from monthly CO2 records at the South Pole. f, Interannual temperature sensitivity of the residual land carbon sink (). The residual land carbon sink of each year is estimated from the CGR by adding the ocean sink and subtracting fossil fuel emission and emission due to land-use change22. In ad, is estimated from Mauna Loa CO2 records. Error bars indicate the 95% confidence interval of derived from 500 bootstrap estimates.

Extended Data Figure 5 Change in drought indices and 20-yr smoothed drought indices over tropical and mid-latitude regions during the past five decades.

a, Change in tropical (23° south to 23° north) annual PDSI, SPEI6 and ORCHIDEE-estimated soil moisture. b, 20-yr smoothed tropical PDSI, SPEI6 and ORCHIDEE-simulated soil moisture. c, Change in mid-latitude (23° north to 48° north) annual PDSI, SPEI6 and ORCHIDEE-simulated soil moisture. d, 20-yr smoothed mid-latitude PDSI, SPEI6 and ORCHIDEE-simulated soil moisture. Years on the x-axis indicate the central year of the 20-yr time window (for example, 1970 represents 1960–1979). All variables are normalized by their respective standard deviations. The changes in SPEI3, SPEI9 and SPEI12 are close to that in SPEI6. Note that SPEI is available till 2011, PDSI is available till 2010 and model soil moisture is available till 2009.

Extended Data Figure 6 Spatial distribution of the difference between the latest and first 20-yr periods during the past five decades in PDSI, SPEI6 and ORCHIDEE-estimated soil moisture.

a, PDSI; b, SPEI6; c, ORCHIDEE-simulated soil moisture. The changes in SPEI3, SPEI9 and SPEI12 are close to that in SPEI6.

Extended Data Figure 7 Change in and the relationship between and precipitation anomalies in the null Monte Carlo experiment.

a, in the first and last 20 yr during 1960–2011. b, for each bin of detrended tropical precipitation anomalies, which are divided into four bins (standardized departure (σ) less than −1, between −1 and 0, between 0 and 1, and greater than 1). Values of for different bins of detrended tropical precipitation anomaly are similar. c, Frequency distribution of the difference between calculated for σ > 1 (wet conditions) and calculated for σ < −1 (dry conditions). The probability of the observed difference (ranging from 5.1 to 6.0 Pg C yr−1 °C−1; Fig. 3) occurring purely by chance is very low (P < 0.01). Error bars indicate the confidence intervals of the corresponding estimates.

Extended Data Figure 8 Change in dCGR/dMAT and and variational effects on estimating dCGR/dMAT between the earlier and latest two decades during 1960–2011.

The effects of interannual variations of tropical precipitation (P), tropical short-wave solar radiation (R), mid-latitude (23° north to 48° north) temperature (MidMAT), mid-latitude precipitation (MidP) and mid-latitude short-wave solar radiation (MidR) are denoted f2 × dP/dMAT, f3 × dR/dMAT, f4 × dMidMAT/dMAT, f5 × dMidP/dMAT and f6 × dMidR/dMAT, respectively, where f2, f3, f4, f5 and f6 are respectively the slopes of P, R, MidMAT, MidP and MidR in the multiple regression of CGR against P, R, MidMAT, MidP, MidR. dy/dx represents the slope of x in the regression of y against x. Error bars indicate the 95% confidence interval of the corresponding value derived from 500 bootstrap estimates.

Extended Data Figure 9 Change in and during 1960–2008 with a 20-yr moving time window.

is the interannual temperature sensitivity of tropical net biome productivity, estimated using five carbon cycle models (HYL, LPJ, ORC, SHE and TRI). To be consistent with model estimated annual net biome productivity, CGR of a specific year is calculated as the difference between the December Mauna Loa CO2 concentration of the year and that of December the previous year. Positive value of indicates reduced anomalies of carbon sinks during warm years.

Extended Data Figure 10 Autocorrelations in CGR and MAT, and their impacts on the estimates of .

a, Autocorrelation coefficients for detrended anomalies of CGR from Mauna Loa during 1960–2011. b, Autocorrelation coefficients for detrended anomalies of MAT during 1960–2011. Dashed lines in a and b indicate 95% confidence bands. c, The comparison between calculated in the multiple regression of interannual variations of the Mauna Loa CGR record against interannual variations in temperature, precipitation and solar radiation (x axis) and calculated in the linear mixed model with same independent variables and a first-order autocorrelation function (y axis). Solid line indicates 1:1 ratio. This 1:1 relationship holds for derived for different time-window lengths.

PowerPoint slides

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Wang, X., Piao, S., Ciais, P. et al. A two-fold increase of carbon cycle sensitivity to tropical temperature variations. Nature 506, 212–215 (2014).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:

This article is cited by


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing