Observations of increased tropical rainfall preceded by air passage over forests

Journal name:
Date published:
Published online
Corrected online

Vegetation affects precipitation patterns by mediating moisture, energy and trace-gas fluxes between the surface and atmosphere1. When forests are replaced by pasture or crops, evapotranspiration of moisture from soil and vegetation is often diminished, leading to reduced atmospheric humidity and potentially suppressing precipitation2, 3. Climate models predict that large-scale tropical deforestation causes reduced regional precipitation4, 5, 6, 7, 8, 9, 10, although the magnitude of the effect is model9, 11 and resolution8 dependent. In contrast, observational studies have linked deforestation to increased precipitation locally12, 13, 14 but have been unable to explore the impact of large-scale deforestation. Here we use satellite remote-sensing data of tropical precipitation and vegetation, combined with simulated atmospheric transport patterns, to assess the pan-tropical effect of forests on tropical rainfall. We find that for more than 60 per cent of the tropical land surface (latitudes 30 degrees south to 30 degrees north), air that has passed over extensive vegetation in the preceding few days produces at least twice as much rain as air that has passed over little vegetation. We demonstrate that this empirical correlation is consistent with evapotranspiration maintaining atmospheric moisture in air that passes over extensive vegetation. We combine these empirical relationships with current trends of Amazonian deforestation to estimate reductions of 12 and 21 per cent in wet-season and dry-season precipitation respectively across the Amazon basin by 2050, due to less-efficient moisture recycling. Our observation-based results complement similar estimates from climate models4, 5, 6, 7, 8, 9, 10, in which the physical mechanisms and feedbacks at work could be explored in more detail.

At a glance


  1. Annual (2001-2007) mean vegetation, precipitation and evaporation.
    Figure 1: Annual (2001–2007) mean vegetation, precipitation and evaporation.

    a, Leaf area index (LAI) from MODIS. b, Example 10-d back-trajectories arriving daily during 2001. Boxes illustrate the four domains analysed in detail in this study. c, Mean cumulative exposure of back-trajectories to LAI over the preceding 10d. d, Precipitation reported by TRMM and other satellites (TRMM3B42). e, Evapotranspiration computed as the mean of the four GLDAS models.

  2. Relationships between daily precipitation and cumulative exposure of 10-d back-trajectories to vegetation LAI (LAI) for 2001-2007.
    Figure 2: Relationships between daily precipitation and cumulative exposure of 10-d back-trajectories to vegetation LAI ( LAI) for 2001–2007.

    a, Plot for air masses arriving in Minas Gerais, Brazil (10–20°S, 40–50°W). Data binned into deciles of LAI and stratified by initial specific humidity (q). Lines show fit to data (solid, wet season; dotted, dry season) and error bars indicate estimation of error in precipitation (Methods Summary). b, Comparison of daily precipitation for air masses that have been exposed to small and large amounts of vegetation (significant (P<0.01) differences indicated by squares at top of panel) during atmospheric transport to the Amazon basin (AF; 10–0°S, 60–70°W), Minas Gerais (MG), the Congo basin (CB; 5°N–5°S, 15–25°E) and south of Congo (ZA; 10–20°S, 20–30°E) (mean, star; median, line; 25th and 75th percentiles, box; 5th and 95th percentiles, whiskers). c, Number of calendar months with significant (P<0.01; red, positive; blue, negative) relationships between precipitation and LAI. Stippling denotes regions where precipitation is a factor of at least two greater in air with large exposure to vegetation than in air with small exposure. Green contour delimits areas with >3m2m−2 annual mean LAI. Black boxes mark the four regions in b.

  3. Atmospheric water-budget components along back-trajectories.
    Figure 3: Atmospheric water-budget components along back-trajectories.

    a, Same as Fig. 2b, but for net change in atmospheric specific humidity (Δq) as a function of LAI. b, Same as Fig. 2b, but for cumulative continental surface evaporation (ΣET; here plotted per day of atmospheric transport) as a function of ΣLAI. c, Same as Fig. 2a, but for precipitation as a function of ET.

  4. Simulated percentage change in precipitation due to 2000-2050 business-as-usual deforestation of the Amazon basin.
    Figure 4: Simulated percentage change in precipitation due to 2000–2050 business-as-usual deforestation of the Amazon basin.

    a, Wet season; b, dry season. Stippling denotes regions where the simulated precipitation anomaly differs from the present-day (1998–2010) rainfall by more than 1s.d. The Amazon (black) and Rio de la Plata (red) basins are marked.

Change history

Corrected online 12 September 2012
Units in Fig. 1c and colour bar values in Fig. 4b were corrected.


  1. Bonan, G. B. Forests and climate change: forcings, feedbacks, and the climate benefit of forests. Science 320, 14441449 (2008)
  2. Shukla, J. & Mintz, Y. Influence of land-surface evapotranspiration on the Earth’s climate. Science 215, 14981501 (1982)
  3. Eltahir, E. A. B. & Bras, R. L. Precipitation recycling in the Amazon basin. Q. J. R. Meteorol. Soc. 120, 861880 (1994)
  4. Henderson-Sellers, A. & Gornitz, V. Possible climatic impacts of land cover transformations, with particular emphasis on tropical deforestation. Clim. Change 6, 231257 (1984)
  5. Lean, J. & Warrilow, D. A. Simulation of the regional climatic impact of Amazon deforestation. Nature 342, 411413 (1989)
  6. Shukla, J., Nobre, C. & Sellers, P. Amazon deforestation and climate change. Science 247, 13221325 (1990)
  7. Werth, D. & Avissar, R. The local and global effects of Amazon deforestation. J. Geophys. Res. 107, 8087 (2002)
  8. Ramos da Silva, R., Werth, D. & Avissar, R. Regional impacts of future land-cover changes on the Amazon basin wet-season climate. J. Clim. 21, 11531170 (2008)
  9. Hasler, N., Werth, D. & Avissar, R. Effects of tropical deforestation on global hydroclimate: a multimodel ensemble analysis. J. Clim. 22, 11241141 (2009)
  10. Nobre, P. et al. Amazon deforestation and climate change in a coupled model simulation. J. Clim. 22, 56865697 (2009)
  11. Pitman, A. J. et al. Uncertainties in climate responses to past land cover change: first results from the LUCID intercomparison study. Geophys. Res. Lett. 36, L14814 (2009)
  12. Butt, N., de Oliveira, P. A. & Costa, M. H. Evidence that deforestation affects the onset of the rainy season in Rondonia, Brazil. J. Geophys. Res. 116, D11120 (2011)
  13. Negri, A. J., Adler, R. F., Xu, L. & Surratt, J. The impact of Amazonian deforestation on dry season rainfall. J. Clim. 17, 13061319 (2004)
  14. Chagnon, F. J. F. & Bras, R. L. Contemporary climate change in the Amazon. Geophys. Res. Lett. 32, L13703 (2005)
  15. Huffman, G. et al. The TRMM multisatellite precipitation analysis (TMPA): quasi-global, multiyear, combined-sensor precipitation estimates at fine scales. J. Hydrometeorol. 8, 3855 (2007)
  16. Myneni, R. B. et al. Global products of vegetation leaf area index and fraction absorbed PAR from year one of MODIS data. Remote Sens. Environ. 83, 214231 (2002)
  17. Gimeno, L., Drumond, A., Nieto, R., Trigo, R. M. & Stohl, A. On the origin of continental precipitation. Geophys. Res. Lett. 37, L13804 (2010)
  18. van der Ent, R. J., Savenije, H. H. G., Schaefli, B. & Steele-Dunne, S. C. Origin and fate of atmospheric moisture over continents. Wat. Resour. Res. 46, W09525 (2010)
  19. Aragão, L. E. O. C. et al. Spatial validation of collection 4 MODIS LAI product in eastern Amazonia. IEEE Trans. Geosci. Rem. Sens. 43, 25262534 (2005)
  20. Rodell, M. et al. The global land data assimilation system. Bull. Am. Meteorol. Soc. 85, 381394 (2004)
  21. Wang, J. F., Bras, R. L. & Eltahir, E. A. B. The impact of observed deforestation on the mesoscale distribution on the mesoscale distribution of rainfall and clouds in Amazonia. J. Hydrometeorol. 1, 267286 (2000)
  22. Garcia-Carreras, L. & Parker, D. J. How does local tropical deforestation affect rainfall? Geophys. Res. Lett. 38, L19802 (2011)
  23. Andreae, M. O. et al. Smoking rain clouds over the Amazon. Science 303, 13371342 (2004)
  24. Koren, I. et al. Aerosol-induced intensification of rain from the tropics to mid-latitudes. Nature Geosci. 5, 118122 (2012)
  25. Soares-Fihlo, B. S. et al. Modelling conservation in the Amazon basin. Nature 440, 520523 (2006)
  26. Phillips, O. L. et al. Drought sensitivity of the Amazon rainforest. Science 323, 13441347 (2009)
  27. Aragão, L. E. O. C. et al. Spatial patterns and fire response of recent Amazonian droughts. Geophys. Res. Lett. 34, L07701 (2007)
  28. Davidson, E. A. et al. The Amazon basin in transition. Nature 481, 321328 (2012)
  29. Huffman, G. J. Estimates of root-mean-square random error for finite samples of estimated precipitation. J. Appl. Meteorol. 36, 11911201 (1997)
  30. Smith, T. M., Arkin, P. A., Bates, J. J. & Huffman, G. J. Estimating bias of satellite precipitation estimates. J. Hydrometeorol. 7, 841856 (2006)
  31. Methven, J. Offline Trajectories: Calculation and Accuracy (Technical Report 44, UK Universities Global Atmospheric Modelling Programme, University of Reading, 1997)
  32. Aragão, L. E. O. C., Shimabukuro, Y. E., Santo, F. D. B. E. & Williams, M. Landscape pattern and spatial variability of leaf area index in Eastern Amazonia. For. Ecol. Mgmt. 211, 240256 (2005)

Download references

Author information


  1. School of Earth and Environment, University of Leeds, Leeds LS2 9JT, UK

    • D. V. Spracklen &
    • S. R. Arnold
  2. Centre for Ecology and Hydrology, Wallingford, Oxfordshire OX10 8BB, UK

    • C. M. Taylor


D.V.S. and S.R.A. initiated the project. All authors participated in discussions, conducted the analysis, assisted with data interpretation and wrote the manuscript.

Competing financial interests

The authors declare no competing financial interests.

Corresponding author

Correspondence to:

Author details

Supplementary information

PDF files

  1. Supplementary Information (555K)

    This file contains Supplementary Tables 1-2 and Supplementary Figures 1-9.

Additional data