Understanding climate change impacts on the Amazon rainforest

January 2013 - The Amazon rainforest is vulnerable through a combination of human influences, including deforestation, fire and climate change. This article examines how the science has moved on since the first Met Office climate model to simulate vegetation response projected extensive Amazon dieback from climate change alone.

As the world's largest store of biodiversity and a major sink for carbon dioxide, the future fate of the Amazon rainforest is of concern to us all. Due to its exceptional biodiversity, around one quarter of all terrestrial species are currently found in the Amazon. The Amazon also represents a major sink for human emissions of carbon dioxide, currently absorbing around 2 billion tons per year. However, this buffer against rising greenhouse gas concentrations is vulnerable, as highlighted by the two major droughts in the past 10 years (in 2005 and 2010).

The 2005 and 2010 droughts

These two droughts (Figure 1: Lewis et al., 2011) both caused substantial carbon loss (Phillips et al., 2009; Lewis et al., 2011), turning the forest from a sink to a temporary source of carbon (if all was converted to carbon dioxide, equivalent to an impact of around 5 billion tons for each event). Some negative effects of the 2005 drought persisted for several years after the event. While the causes of these events are not well understood, they demonstrate that substantial drought is possible, and that the forest is sensitive to such rainfall deficits. Other recent work has confirmed that global forest species tend to have relatively narrow tolerance of drought (Choat et al., 2012). A concern exists that such droughts could become more common in the future, through climate change, but the fate of the Amazon is an even more complex question.

Rainfall deficits in the 2005 and 2010 Amazon droughts (From Lewis et al., 2011). Figure 1. Rainfall deficits in the 2005 and 2010 Amazon droughts (From Lewis et al., 2011). (A and B) Satellite-derived standardized anomalies for dry-season rainfall for the two most extensive droughts of the 21st century in Amazonia. (C and D) The difference in the 12-month (October to September) Maximum climatological water deficit (mm) from the decadal mean (excluding 2005 and 2010), a measure of drought intensity that correlates with tree mortality. (A) and (C) show the 2005 drought

The combined influence of deforestation, fire and climate change

The first Met Office global climate model to include the response of vegetation had shown the possibility of extensive forest loss from one factor: climate change from increasing greenhouse gases alone (Cox et al. 2000, 2004). Further work at the Met Office and elsewhere showed that the full set of IPCC AR4 climate models showed agreement in projecting a general long-term drying trend in the Amazon (Malhi, Roberts, Betts et al, 2007) although the HadCM3LC results were at the extreme end of the range (Betts et al, 2012). One of the reasons for the extreme result in HadCM3LC was the existence of strong feedbacks arising from the forest loss itself (Betts et al, 2004). While this result highlighted the potential vulnerability of this critical ecosystem, the scientific debate has evolved to consider the combined threats from deforestation, fire and climate change (Betts et al, 2008). As well as having substantial individual effects, these factors may combine in complex ways. For example, deforestation in one part of the Amazon can reduce rainfall elsewhere in the forest, potentially making it more vulnerable to dieback from climate change. Fire frequency and intensity is affected by both human activity and climate, with consequences for forest. The forest density may in turn affect fire activity by suppressing grass fuels, a feedback which can in principle allow abrupt change between forest and savannah.

The Met Office is collaborating with international colleagues, for example in the AMAZALERT and EMBRACE projects, to understand and account for these processes in policy advice.

Amazon vulnerability to climate change from increased carbon dioxide

Even considering just the effect of future climate change from increasing carbon dioxide, the range of possible outcomes for the Amazon is large. A broad range of climate changes are projected, associated in part with uncertainty in how the tropical Pacific and Atlantic sea-surface temperatures will respond in the future and how this will influence regional rainfall patterns (Good et al., 2008). However, climate models do show greater agreement on the risk of future drought in the eastern part of the Amazon.

Projections from the new Met Office model, HadGEM2-ES

This scientific uncertainty is illustrated by the difference in projections from two generations of Met Office global climate models (Good et al., 2013, Figure 2). While the previous generation model had projected extensive dieback from increasing carbon dioxide alone, the new Met Office model, HadGEM2-ES, shows little forest change by the end of the century (although some forest loss is seen as the trees respond over longer time periods). This difference between the two models arises from major structural changes, taking advantage of greater scientific understanding, and increased supercomputing power. A combination of factors affect the differing forest futures: present climate, climate change and forest sensitivity. Although a range of factors are likely to have contributed, the difference between the model responses is mainly attributable to the length of dry season, which is partly driven by changes in Atlantic sea surface temperature patterns.

Future forest fraction old and new models Figure 2. Future forest fraction under an idealised scenario of steadily increasing carbon dioxide (roughly corresponding to the end of the century under a business as usual emissions scenario), from the old (top) and new (bottom) Met Office models. These results exclude the effects of deforestation and fire. Some forest dieback is found in the new model as forest adjusts over longer timescales.

It is not surprising that such a significantly improved model should differ in this way from its predecessor. Although it is now less likely, the earlier projection of extensive Amazon dieback from increasing carbon dioxide alone (Cox et al., 2000, 2004) nevertheless remains a possible outcome. No new scientific evidence has emerged to rule this possibility out, although the fact that such severe climate change has not been reproduced by other climate models suggests that it is not the most likely outcome. However, climate models are in greater agreement on the risk of future drought in the eastern part of the Amazon. When the complex effects of deforestation and fire are accounted for (not included in the Met Office projections described above, Good et al., 2013), the risk of substantial forest loss remains of major concern.

In its advice to policy makers, the Met Office takes account of various lines of evidence, including observations (such as the carbon losses during the 2005 and 2010 droughts) and results from a range of international climate models. The advice on the potential vulnerability of the Amazon has not substantially changed following the introduction of the new Met Office model.

Advancing our understanding

Science is always advancing, new knowledge is always being gained, and news results are always emerging. This is true of all sciences and climate science is no different. As future increases in computing power and scientific understanding allow more detailed and accurate representation of the vast and complex range of processes affecting the Amazon forest, so our assessments of potential vulnerability will change and become more robust. For example, more powerful supercomputers will permit finer spatial resolution, which may lead to significant change in tropical Atlantic sea-surface temperature projections. Full consideration of the vast range of different tree species and how they individually respond to stressors is a further area of research. Furthermore, new understanding of the nitrogen cycle, deforestation and fire could significantly affect the assessed future resilience of tropical forest.

This latest study is one piece of evidence that helps to inform our understanding of the Earth's response to climate change. Along with observational studies and a wide range of other inputs, this informs the consensus of our scientific advice. Our fundamental advice to policymakers on the risks of global warming remains unchanged. While Amazon die-back remains a low probability, but high impact potential consequence of our changing climate, an expectation of less extreme changes in the length of the Amazonian dry season means that die-back of the Amazon rainforest as a result of climate change alone is now considered less likely.  

References

  • Betts, R.A., P. M. Cox, M. Collins, P. P. Harris, C. Huntingford, and C. D. Jones (2004). The role of ecosystem-atmosphere interactions in simulated Amazonian precipitation decrease and forest dieback under global climate warming. Theor. Appl. Climatol., 78:157-175, 2004.
  • Richard A. Betts, Yadvinder Malhi, and J. Timmons Roberts (2008) The future of the Amazon: new perspectives from climate, ecosystem and social sciences. Phil. Trans. R. Soc. 363 (1498), 1729-1735
  • Richard A. Betts, Nigel W. Arnell, Penelope Boorman, Sarah E. Cornell, Joanna I. House, Neil R. Kaye, Mark P. McCarthy, Doug McNeall, Michael G. Sanderson and Andrew J. Wiltshire (2012) Climate change impacts and adaptation: an Earth system view. In: Cornell, S., Prentice, C., House, J. and Downy, C. (Eds), Understanding the Earth System: Global Change Science for Application. Cambridge University Press
  • Choat, B., et al. (2012), Global convergence in the vulnerability of forests to drought, Nature, 491(7426), 752.
  • Cox, P. M., R. A. Betts, C. D. Jones, S. A. Spall, and I. J. Totterdell (2000), Acceleration of global warming due to carbon-cycle feedbacks in a coupled climate model (vol 408, pg 184, 2000), Nature, 408(6813), 750-750.
  • Cox, P. M., R. A. Betts, M. Collins, P. P. Harris, C. Huntingford, and C. D. Jones (2004), Amazonian forest dieback under climate-carbon cycle projections for the 21st century, Theor Appl Climatol, 78(1-3), 137-156.
  • Good, P., J. A. Lowe, M. Collins, and W. Moufouma-Okia (2008), An objective tropical Atlantic sea surface temperature gradient index for studies of south Amazon dry-season climate variability and change, Philos T R Soc B, 363(1498), 1761-1766.
  • Good, P., C. D. Jones, J. A. Lowe, R. A. Betts, and N. Gedney (2013), Comparing tropical forest projections from two generations of Hadley Centre Earth System models, HadGEM2-ES and HadCM3LC, Journal of Climate, 26 (2), 495-511
  • Lewis, S. L., P. Brando, O. L. Phillips, G. Van der Heijden, and D. Nepstad (2011), The 2010 Amazon Drought, Science, 331(6017).
  • Malhi, Y., L. E. O. C. Aragao, D. Galbraith, C. Huntingford, R. Fisher, P. Zelazowski, S. Sitch, C. McSweeney, and P. Meir (2009), Exploring the likelihood and mechanism of a climate-change-induced dieback of the Amazon rainforest, Proceedings of the National Academy of Sciences of the United States of America, 106(49), 20610-20615.
  • Phillips, O. L., et al. (2009), Drought Sensitivity of the Amazon Rainforest, Science, 323(5919), 1344-1347. 

Last updated: 14 August 2014