College of Engineering | Apply to CSU | Disclaimer | Equal Opportunity Statement | Privacy | Search CSU

The Impacts of Amazon Deforestation on Pacific Climate

You must be on the CSU network—either physically or using VPN—to watch this or any of the videos on this site.

July 25, 2016
Leah Lindsey
Hosted by Dave Randall (advisor), Scott Denning, Allan Kirkpatrick (Mechanical Engineering)


Variability in eastern Pacific sea surface temperatures (SSTs) associated with the El Niño Southern Oscillation are known to affect Amazonian precipitation, but to what extent do changing Amazonian vegetation and rainfall impact eastern Pacific SST? The Amazon rainforest is threatened by many factors including climate change and clearing for agricultural reasons. Forest fires and dieback are more likely due to increased frequency and intensity of droughts in the region. It is possible that extensive Amazon deforestation can enhance El Niño conditions by weakening the Walker circulation.

Correlations between annual rainfall rates over the Amazon and other atmospheric parameters (global precipitation, surface air temperature, low cloud amount, 500 hPa vertical velocity, surface winds, and 200 hPa winds) over the eastern Pacific indicate strong relationships among these fields. Maps of these correlations (teleconnection maps) reveal that when the Amazon is rainy SSTs in the central and eastern Pacific are cold, rainfall is suppressed over the central and eastern Pacific, low clouds are prominent over the eastern and southeastern Pacific, and subsidence over the central and eastern Pacific is enhanced. Precipitation in the Amazon is also consistent with a strong Walker circulation (La Niña conditions), manifest as strong correlations with the easterly surface and westerly 200 hPa zonal winds. Coupling between Amazon rainfall and these fields are seen in observations and model data. Correlations were calculated using data from observations, reanalysis data, two models under the Coupled Model Intercomparison Project/Atmospheric Model Intercomparison Project (CMIP5/AMIP), and an AMIP run with the model used in this study, the Community Earth System Model (CESM1.1.1). Although the correlations between Amazon precipitation and the aforementioned fields are strong, they do not show causality. In order to investigate the impact of tropical South American deforestation on the Pacific climate, numerical experiments were performed using the CESM.

Amazon deforestation was studied in an idealized world where a single continent was covered in forest and then, in a separate simulation, covered in grassland. Four different sets of simulations were carried out: 1) the baseline idealized set-up with prescribed SST, 2) another with an Andes-like mountain range, 3) a simulation with a slab ocean model rather than prescribed SST, and 4) a simulation repeated with the standard Community Atmosphere Model (CAM4) replaced by the Superparameterized version (SP-CAM). The continent in these simulations was compared to the Amazon, and the ocean to the west of the continent was compared to the eastern Pacific.

All of the simulations showed a strong warming of around 3-4°C over the continent going from forest to grassland. A notable decrease in precipitation over land of about 1-3 mm day-1 and increase to the west of the continent of about 1-2 mm day-1 was also observed in most of the simulations. The simulations with the slab ocean model showed enhanced precipitation changes with a corresponding decrease of 2-4 mm day-1 over land and increase of 3-5 mm day-1 west of the continent. Simulations that used the SP-CAM showed very small changes in precipitation, which was likely due to the decreased spin-up time allowed for these simulations. The decrease in the surface roughness and reduction in the evapotranspiration for the simulations with grassland contributed to these changes in surface temperature and precipitation. The conversion of forest to grassland in our experiments imply that deforestation can lead to weakening of the Walker circulation by weakening easterly surface winds and westerly upper tropospheric winds. These findings suggest that large-scale Amazon deforestation is capable of enhancing El Niño conditions.