Letter | Published:

Sensitivity of atmospheric CO2 growth rate to observed changes in terrestrial water storage


Land ecosystems absorb on average 30 per cent of anthropogenic carbon dioxide (CO2) emissions, thereby slowing the increase of CO2 concentration in the atmosphere1. Year-to-year variations in the atmospheric CO2 growth rate are mostly due to fluctuating carbon uptake by land ecosystems1. The sensitivity of these fluctuations to changes in tropical temperature has been well documented2,3,4,5,6, but identifying the role of global water availability has proved to be elusive. So far, the only usable proxies for water availability have been time-lagged precipitation anomalies and drought indices3,4,5, owing to a lack of direct observations. Here, we use recent observations of terrestrial water storage changes derived from satellite gravimetry7 to investigate terrestrial water effects on carbon cycle variability at global to regional scales. We show that the CO2 growth rate is strongly sensitive to observed changes in terrestrial water storage, drier years being associated with faster atmospheric CO2 growth. We demonstrate that this global relationship is independent of known temperature effects and is underestimated in current carbon cycle models. Our results indicate that interannual fluctuations in terrestrial water storage strongly affect the terrestrial carbon sink and highlight the importance of the interactions between the water and carbon cycles.

Access optionsAccess options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.


  1. 1.

    Le Quéré, C. et al. Global carbon budget 2017. Earth Syst. Sci. Data 10, 405–448 (2018).

  2. 2.

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

  3. 3.

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

  4. 4.

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

  5. 5.

    Wang, J., Zeng, N. & Wang, M. R. Interannual variability of the atmospheric CO2 growth rate: roles of precipitation and temperature. Biogeosciences 13, 2339–2352 (2016).

  6. 6.

    Fang, Y. et al. Global land carbon sink response to temperature and precipitation varies with ENSO phase. Environ. Res. Lett. 12, 064007 (2017).

  7. 7.

    Tapley, B. D., Bettadpur, S., Ries, J. C., Thompson, P. F. & Watkins, M. M. GRACE measurements of mass variability in the Earth System. Science 305, 503–505 (2004).

  8. 8.

    Peters, G. P. et al. Towards real-time verification of CO2 emissions. Nat. Clim. Chang. 7, 848–850 (2017).

  9. 9.

    Sitch, S. et al. Recent trends and drivers of regional sources and sinks of carbon dioxide. Biogeosciences 12, 653–679 (2015).

  10. 10.

    Friedlingstein, P. et al. Uncertainties in CMIP5 climate projections due to carbon cycle feedbacks. J. Clim. 27, 511–526 (2014).

  11. 11.

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

  12. 12.

    Ciais, P. et al. Europe-wide reduction in primary productivity caused by the heat and drought in 2003. Nature 437, 529–533 (2005).

  13. 13.

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

  14. 14.

    Beer, C. et al. Terrestrial gross carbon dioxide uptake: global distribution and covariation with climate. Science 329, 834–838 (2010).

  15. 15.

    Humphrey, V., Gudmundsson, L. & Seneviratne, S. I. A global reconstruction of climate-driven subdecadal water storage variability. Geophys. Res. Lett. 44, 2300–2309 (2017).

  16. 16.

    Lucht, W. et al. Climatic control of the high-latitude vegetation greening trend and Pinatubo effect. Science 296, 1687–1689 (2002).

  17. 17.

    Trenberth, K. E. & Dai, A. Effects of Mount Pinatubo volcanic eruption on the hydrological cycle as an analog of geoengineering. Geophys. Res. Lett. 34, (2007).

  18. 18.

    Mercado, L. M. et al. Impact of changes in diffuse radiation on the global land carbon sink. Nature 458, 1014–1017 (2009).

  19. 19.

    Döll, P., Müller Schmied, H., Schuh, C., Portmann, F. T. & Eicker, A. Global-scale assessment of groundwater depletion and related groundwater abstractions: combining hydrological modeling with information from well observations and GRACE satellites. Wat. Resour. Res. 50, 5698–5720 (2014).

  20. 20.

    Rodell, M. et al. The global land data assimilation system. Bull. Am. Meteorol. Soc. 85, 381–394 (2004).

  21. 21.

    Ni, S. et al. Global terrestrial water storage changes and connections to ENSO events. Surv. Geophys. 39, 1–22 (2018).

  22. 22.

    Jung, M. et al. Compensatory water effects link yearly global land CO2 sink changes to temperature. Nature 541, 516–520 (2017).

  23. 23.

    Tramontana, G. et al. Predicting carbon dioxide and energy fluxes across global FLUXNET sites with regression algorithms. Biogeosciences 13, 4291–4313 (2016).

  24. 24.

    Fan, Y., Miguez-Macho, G., Jobbágy, E. G., Jackson, R. B. & Otero-Casal, C. Hydrologic regulation of plant rooting depth. Proc. Natl Acad. Sci. USA 114, 10572–10577 (2017).

  25. 25.

    Battin, T. J. et al. The boundless carbon cycle. Nat. Geosci. 2, 598–600 (2009).

  26. 26.

    Poulter, B. et al. Contribution of semi-arid ecosystems to interannual variability of the global carbon cycle. Nature 509, 600–603 (2014).

  27. 27.

    Ahlstrom, A. et al. The dominant role of semi-arid ecosystems in the trend and variability of the land CO2 sink. Science 348, 895–899 (2015).

  28. 28.

    Orlowsky, B. & Seneviratne, S. I. Elusive drought: uncertainty in observed trends and short- and long-term CMIP5 projections. Hydrol. Earth Syst. Sci. 17, 1765–1781 (2013).

  29. 29.

    Scanlon, B. R. et al. Global models underestimate large decadal declining and rising water storage trends relative to GRACE satellite data. Proc. Natl Acad. Sci. USA 115, E1080–E1089 (2018).

  30. 30.

    Swann, A. L. S., Hoffman, F. M., Koven, C. D. & Randerson, J. T. Plant responses to increasing CO2 reduce estimates of climate impacts on drought severity. Proc. Natl Acad. Sci. USA 113, 10019–10024 (2016).

  31. 31.

    Wahr, J., Swenson, S., Zlotnicki, V. & Velicogna, I. Time-variable gravity from GRACE: first results. Geophys. Res. Lett. 31, L11501 (2004).

  32. 32.

    Wahr, J., Molenaar, M. & Bryan, F. Time variability of the Earth’s gravity field: hydrological and oceanic effects and their possible detection using GRACE. J. Geophys. Res. Solid Earth 103, 30205–30229 (1998).

  33. 33.

    Wouters, B. et al. GRACE, time-varying gravity, Earth system dynamics and climate change. Rep. Prog. Phys. 77, (2014).

  34. 34.

    Cazenave, A. et al. The rate of sea-level rise. Nat. Clim. Chang. 4, 358–361 (2014).

  35. 35.

    Adhikari, S. & Ivins, E. R. Climate-driven polar motion: 2003-2015. Sci. Adv. 2, (2016).

  36. 36.

    Abelen, S. & Seitz, F. Relating satellite gravimetry data to global soil moisture products via data harmonization and correlation analysis. Remote Sens. Environ. 136, 89–98 (2013). 

  37. 37.

    Humphrey, V., Gudmundsson, L. & Seneviratne, S. I. Assessing global water storage variability from GRACE: trends, seasonal cycle, subseasonal anomalies and extremes. Surv. Geophys. 37, 357–395 (2016).

  38. 38.

    Wiese, D. N., Landerer, F. W. & Watkins, M. M. Quantifying and reducing leakage errors in the JPL RL05M GRACE mascon solution. Wat. Resour. Res. 52, 7490–7502 (2016).

  39. 39.

    Watkins, M. M., Wiese, D. N., Yuan, D. N., Boening, C. & Landerer, F. W. Improved methods for observing Earth’s time variable mass distribution with GRACE using spherical cap mascons. J. Geophys. Res. Solid Earth 120, 2648–2671 (2015).

  40. 40.

    Scanlon, B. R. et al. Global evaluation of new GRACE mascon products for hydrologic applications. Wat. Resour. Res. 52, 9412–9429 (2016).

  41. 41.

    Masarie, K. A. & Tans, P. P. Extension and integration of atmospheric carbon-dioxide data into a globally consistent measurement record. J. Geophys. Res. D 100, 11593–11610 (1995).

  42. 42.

    Dlugokencky, E. & Tans, P. Trends in Atmospheric Carbon Dioxide. http://www.esrl.noaa.gov/gmd/ccgg/trends/ (National Oceanic and Atmospheric Administration, Earth System Research Laboratory (NOAA/ESRL), 2014).

  43. 43.

    Beck, H. E. et al. MSWEP: 3-hourly 0.25 degrees global gridded precipitation (1979-2015) by merging gauge, satellite, and reanalysis data. Hydrol. Earth Syst. Sci. 21, 589–615 (2017).

  44. 44.

    Adler, R. F. et al. The version-2 Global Precipitation Climatology Project (GPCP) monthly precipitation analysis (1979-present). J. Hydrometeorol. 4, 1147–1167 (2003).

  45. 45.

    Huffman, G. J., Adler, R. F., Bolvin, D. T. & Gu, G. J. Improving the global precipitation record: GPCP version 2.1. Geophys. Res. Lett. 36, (2009).

  46. 46.

    Harris, I., Jones, P. D., Osborn, T. J. & Lister, D. H. Updated high-resolution grids of monthly climatic observations—the CRU TS3.10 dataset. Int. J. Climatol. 34, 623–642 (2014).

  47. 47.

    Rohde, R. et al. A new estimate of the average Earth surface land temperature spanning 1753 to 2011. Geoinform. Geostatist. 01, http://doi.org/10.4172/2327-4581.1000101 (2013).

  48. 48.

    Wolter, K. & Timlin, M. S. Measuring the strength of ENSO events: how does 1997/98 rank? Weather 53, 315–324 (1998).

  49. 49.

    Friedl, M. A. et al. MODIS Collection 5 global land cover: algorithm refinements and characterization of new datasets. Remote Sens. Environ. 114, 168–182 (2010).

  50. 50.

    Mudelsee, M. Climate Time Series Analysis: Classical Statistical and Bootstrap Methods 2nd edn, Chs 4 and 7 (Springer, Cham, 2014).

Download references


All datasets supporting the results of this paper are openly accessible from the references listed in Supplementary Table 1. This research was funded by the European Research Council DROUGHT-HEAT project (contract 617518). P.C. was supported by the European Research Council Synergy grant ERC-2013-SyG-610028 IMBALANCE-P. We thank M. Jung and U. Weber for providing the water availability index used in FluxCom and R. Wartenburger for technical support. We gratefully thank the following data providers and model developers for their continuous efforts and for sharing their data: the NASA Jet Propulsion Laboratory, the NOAA Earth System Research Laboratory, the Global Carbon Project, the WaterGAP Global Hydrology Model (WGHM), the Global Land Data Assimilation System (GLDAS), Multi-Source Weighted-Ensemble Precipitation (MSWEP), the Global Precipitation Climatology Project (GPCP), the University of East Anglia Climatic Research Unit (CRU), Berkeley Earth, and all contributors as well as data providers to the FluxCom initiative and the TRENDY experiment version 3, which included the models CABLE, CLM, ISAM, JSBACH, JULES, LPJ, LJP-GUESS, LPX-Bern, ORCHIDEE, VEGAS and VISIT.

Reviewer information

Nature thanks A. Dolman, C. Funk and B. Zaitchik for their contribution to the peer review of this work.

Author information

V.H., S.I.S., J.Z. and P.C. designed the study. V.H. conducted the data analysis with support from L.G., J.Z., S.S. and S.I.S., and wrote the manuscript. The interpretation, final text and figures resulted from the contributions of all co-authors.

Competing interests

: The authors declare no competing interests.

Correspondence to Vincent Humphrey or Sonia I. Seneviratne.

Extended data figures and tables

Extended Data Fig. 1 Ecosystems respond to water storage.

Water storage is more relevant than precipitation when investigating the impacts of changes in water availability on ecosystems.

Extended Data Fig. 2 Reproduction of Fig. 1c, d with GRACE-REC.

Composite mean TWS anomalies associated with the 5% highest (a) and 5% lowest (b) monthly CGR (n = 20 months in each case) based on GRACE-REC (that is, covering the 1980–2016 time period). Inset bar plots indicate the season of the months used in the composite.

Extended Data Fig. 3 ENSO, precipitation and TWS.

Because it integrates precipitation anomalies, water storage is slightly phase-shifted with respect to ENSO and precipitation time series. Here, El Niño (La Niña) conditions correspond to the periods where the Multivariate ENSO Index (MEI) exceeds 0.5 (−0.5). The strongest ENSO events (MEI > 1 or MEI < −1) are shown in darker colours.

Extended Data Fig. 4 Dominant contribution of precipitation to TWS anomalies.

a, Global means of GRACE, GRACE-REC and GRACE-REC driven only with precipitation anomalies. The statistical reconstruction of GRACE (GRACE-REC) is calibrated with both precipitation and temperature information15. We use this model to predict the precipitation-driven component of the TWS signal (by setting temperature variability to zero). Most of the global TWS signal can be reconstructed based on precipitation anomalies only. b, Performance of the GRACE-REC model at the grid scale. c, Contribution of precipitation to the locally reconstructed TWS. A comparison between GRACE-REC, global hydrological models and GRACE can also be found in ref. 15.

Extended Data Fig. 5 Reproduction of Fig. 3 with mean precipitation.

Same as Fig. 3, but using yearly precipitation P from the Global Precipitation Climatology Project (with a 4-month lag) instead of TWS from GRACE. Significance (P < 0.05, n = 15; Methods) is indicated with crosses.

Extended Data Fig. 6 Illustration of soil moisture signals with different fractions of IAV.

The fraction of IAV quantifies the importance of low frequency variability in the overall variance of a given signal. Here, it is defined as the ratio between the variance of the yearly (de-trended) time series (b) and the variance of the monthly anomalies (a) (see Methods). The fraction of IAV tends to increase when deeper soil layers are included (c). This is because deeper layers have a longer residence time (or memory) and thus respond more slowly to changes in the meteorological forcing. Illustrative data based on GLDAS2-Noah, extracted for Spain (4.25° W, 40.25° N).

Extended Data Fig. 7 Fraction of IAV in water storage changes.

a, Average fraction of IAV in water storage changes simulated by DGVMs and FluxCom (which typically only include root-zone soil moisture). b, Fraction of IAV in water storage changes observed by GRACE (which include all water reservoirs). To ensure comparability between models and GRACE, model outputs were first averaged to the spatial resolution of GRACE. We note that unlike modelled soil moisture, GRACE observations suffer from measurement errors that tend to increase the high-frequency (month-to-month) variability. Therefore, the fraction of IAV retrieved from GRACE would be even higher if there were no measurement errors in GRACE.

Extended Data Fig. 8 Distribution of the fraction of IAV by land cover classes.

This compares the values shown in the maps of Extended Data Fig. 7 for different land-cover classes. The fraction of IAV found in GRACE TWS (dark blue) is higher compared to models (green). Because GRACE observations are contaminated by high-frequency measurement errors, the fraction of IAV found in GRACE is shifted towards lower values. Here, the fraction of IAV derived from GRACE-REC (light blue) may provide a more robust estimate of the actual fraction of IAV in TWS. Adding GRACE measurement errors (as provided with GRACE NASA-JPL data) to the GRACE-REC data reproduces very well the overall shift (dashed light blue) towards lower values that occurs with original GRACE data.

Extended Data Fig. 9 Relationship between the fraction of IAV in model water storage and the fraction of IAV in NEEwater.

a, c, Mean fraction of IAV obtained at all grid cells for TRENDY (a) and FluxCom (c), with point cloud density indicated by the colour shading. The fraction of IAV in NEEwater is directly limited by the fraction of IAV present in the underlying water storage signal. b, d, The same as a and c, stratified by land-cover class. In land-cover classes that are typically moisture-limited (for example, semi-arid), the fraction of IAV in NEEwater is potentially strongly limited by the fraction of IAV in water storage. e, This relationship is also found for the global mean signals of the individual models.

Extended Data Fig. 10 Contribution of six different land cover types to the global water storage signal.

a, GRACE TWS anomalies by land-cover type, smoothed with a 6-month moving average and offset for readability. b, Regional contributions to the global water storage signal. High values indicate that a region bears a high contribution to the overall global mean water storage signal. This metric is based on the definition proposed in ref. 27 for analysing regional contributions to global net biome production. The value reported for the models is the mean across all models.

Supplementary information

Supplementary Information

This file contains Supplementary Tables 1-3 with links to all datasets used in this analysis. It also contains Supplementary Figures 1-10 which either complement minor aspects of the main text or reproduce the main figures with slightly different approaches.

Source data

Source Data Fig. 1

Source Data Fig. 2

Source Data Fig. 3

Source Data Fig. 4

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Further reading

Fig. 1: IAV in CGR and TWS.
Fig. 2: Correlations between CGR and meteorological drivers over different spatial domains at monthly and yearly scale.
Fig. 3: Confounding effects of water storage and temperature on correlations with CGR.
Fig. 4: Observed and modelled relations between global water storage, temperature and carbon fluxes.
Extended Data Fig. 1: Ecosystems respond to water storage.
Extended Data Fig. 2: Reproduction of Fig. 1c, d with GRACE-REC.
Extended Data Fig. 3: ENSO, precipitation and TWS.
Extended Data Fig. 4: Dominant contribution of precipitation to TWS anomalies.
Extended Data Fig. 5: Reproduction of Fig. 3 with mean precipitation.
Extended Data Fig. 6: Illustration of soil moisture signals with different fractions of IAV.
Extended Data Fig. 7: Fraction of IAV in water storage changes.
Extended Data Fig. 8: Distribution of the fraction of IAV by land cover classes.
Extended Data Fig. 9: Relationship between the fraction of IAV in model water storage and the fraction of IAV in NEEwater.
Extended Data Fig. 10: Contribution of six different land cover types to the global water storage signal.


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.