Nonlinear interactions between the Amazon rainforest and the atmospheric moisture circulation in South America: Possible impacts of ongoing deforestation
Statistical Physics, Nonlinear Dynamics, Complex Systems, Soft and Fluid Matter, Biological Physics
Final Report Abstract
The aim of this project was to investigate the impacts of deforestation of the Amazon rainforest on the regional atmospheric moisture circulation patterns over South America, with main focus on a positive feedback between atmospheric latent heating over the Amazon basin and the moisture inflow from the Atlantic ocean. The key research question was whether this feedback could be jeopardized by ongoing deforestation via reductions of evapotranspiration, resulting in reduced net available moisture amounts, and hence reduced latent heating. Furthermore, it was envisioned to study the impacts of a collapse of this feedback on precipitation in the western Amazon basin, and further downstream toward subtropical South America. The project was separated into three different steps: First, the relevance of the latent-heating related feedback was studied on the basis of a onedimensional model based on the moisture balance equations. This model was first evaluated with respect to present-day observations of relevant atmospheric and hydrological variables, and then used to simulate the impacts of successive deforestation. The simulation results suggest that there exists a ‘tipping point’ in the rate of deforestation, beyond which precipitation in the western Amazon basin and further downstream abruptly reduces by about 40%. This abrupt reduction is associated with a double-fold bifurcation of the one-dimensional model, indicating that the consequences of crossing the critical deforestation rates cannot easily be reverted by subsequent reforestation. In the second step, the one-dimensional model should have been generalized to two horizontal spatial dimensions using complex networks. Unfortunately, however, during the time span between the proposal submission and the final start of the project, two studies we published with very similar purpose. Moreover, I noticed in course that the network constructions in these studies, as well as my own initial ideas to construct a two-dimensional network to represent the moisture flow across the Amazon basin, including cascading effects caused by successive precipitation and evapotranspiration, were flawed by the fact that these representations cannot account for changes in the directions of the low-level wind field resulting from changing latent heating patterns caused by widespread deforestation. In a third step, the impacts of deforestation on regional climate in South America were supposed to be studied using a regional climate model. Several simulation studies with similar topics were published between proposal submission and final start date of the project; in particular, a meta-analysis of 96 relevant studies indicates that model uncertainty, and in particular the parameter uncertainty associated with coupling atmospheric and vegetation processes, are currently so high that different models strongly diverge in their simulated impacts of Amazon deforestation. Therefore, I decided to refrain from using these models, and to focus instead on further investigating the resilience of the Amazon basin, as well as the potential, global-scale impacts of a degradation of this ecosystem, on a purely observational basis. The main results of this final part of the project are that 1. the resilience of the Amazon rainforest against reducing precipitation amounts – as projected for ongoing global warming scenarios – is significantly enhanced in regions with relatively higher inter-annual rainfall variability during preceding decades, 2. droughts in the Amazon, which are projected to increase in frequency and duration, can be predicted using tropical Atlantic sea-surface temperature anomalies, and 3. there exist significant atmospheric teleconnections between the Amazon basin on the one hand, and the United States, eastern Europe and the Mediterranean, as well as southern Africa, on the other hand, which are hence likely to be impacted by a degradation of the Amazon ecosystem.
Publications
- A deforestation-induced tipping point for the South American monsoon system, Nature Scientific Reports 7, 41489 (2017)
N. Boers, N. Marwan, H.M.J. Barbosa, J. Kurths
(See online at https://doi.org/10.1038/srep41489) - Higher resilience to climatic disturbances in tropical vegetation exposed to more variable rainfall, Nature Geoscience, 2019
C. Ciemer, N. Boers, M. Hirota, J. Kurths, F. Müller-Hansen, R. Oliveira, R. Winkelmann
(See online at https://doi.org/10.1038/s41561-019-0312-z)
