期刊名称：Proceedings of the National Academy of Sciences
印刷版ISSN：0027-8424
电子版ISSN：1091-6490
出版年度：2015
卷号：112
期号：43
页码：13172-13177
DOI：10.1073/pnas.1421010112
语种：English
出版社：The National Academy of Sciences of the United States of America
摘要：SignificanceRecent severe droughts in the Amazon basin have increased interest in future climatological and ecological conditions of this region. Future changes in drought and wet periods could have enormous impacts on forest structure, biomass, and composition, but our ability to predict changes in the hydrological regime remains highly uncertain. We evaluate an ensemble of state-of-the-art climate models and demonstrate their accuracy in simulating processes influencing drought in Amazonia. These models provide projections of future properties of drought and wet periods, and indicate that different portions of the Amazon Basin undergo contrasting hydrological futures but will share a tendency toward more hydrological extremes. Future intensification of Amazon drought resulting from climate change may cause increased fire activity, tree mortality, and emissions of carbon to the atmosphere across large areas of Amazonia. To provide a basis for addressing these issues, we examine properties of recent and future meteorological droughts in the Amazon in 35 climate models participating in the Coupled Model Intercomparison Project Phase 5 (CMIP5). We find that the CMIP5 climate models, as a group, simulate important properties of historical meteorological droughts in the Amazon. In addition, this group of models reproduces observed relationships between Amazon precipitation and regional sea surface temperature anomalies in the tropical Pacific and the North Atlantic oceans. Assuming the Representative Concentration Pathway 8.5 scenario for future drivers of climate change, the models project increases in the frequency and geographic extent of meteorological drought in the eastern Amazon, and the opposite in the West. For the region as a whole, the CMIP5 models suggest that the area affected by mild and severe meteorological drought will nearly double and triple, respectively, by 2100. Extremes of wetness are also projected to increase after 2040. Specifically, the frequency of periods of unusual wetness and the area affected by unusual wetness are projected to increase after 2040 in the Amazon as a whole, including in locations where annual mean precipitation is projected to decrease. Our analyses suggest that continued emissions of greenhouse gases will increase the likelihood of extreme events that have been shown to alter and degrade Amazonian forests.
