Skip to Main Content U.S. Department of Energy
Science Directorate
Page 6 of 740

Atmospheric Sciences & Global Change
Research Highlights

December 2018

Restoring Rain to a Model Amazon Forest

Simulations reveal terrestrial factors and airflow as contributors to climate model challenges in producing rainfall over the world's largest rainforest.

topography of the Amazon and its coastlines affect airflow near the surface
The complex topography of the Amazon and its coastlines affect airflow near the surface, deflecting winds that, in turn, control where water vapor converges to produce clouds and rain. Enlarge Image.

The Science

Most global climate models are unable to accurately simulate rainfall over the Amazon forest, but the reasons are not well understood. To investigate this question, scientists at the U.S. Department of Energy's Pacific Northwest National Laboratory analyzed more than 20 different simulations from Community Earth System Model version 1 (CESM1). They found that the complex topography and coastlines surrounding the Amazon change the airflow that influences water vapor, temperature, and convection in the region. By modifying the model to more directly reflect the effects of airflow on convection, the team found they could more closely simulate the observed Amazon rainfall.

The Impact

The vast Amazon forest has significant influence on global climate, cycling large amounts of water and carbon between the atmosphere and the terrestrial system. Forest productivity depends in large part on rainfall, and the limitations of climate models in reflecting the Amazon rainfall has left a large uncertainty in projecting future changes of the forest as well as global climate. Identifying factors that affect airflow in the Amazon helps researchers improve models of rainfall for the region and builds greater confidence in understanding future changes of global climate.


Abundant rainfall over the Amazon sustains the world's largest tropical forest that plays a significant role in the global-scale water and carbon cycles. But the Amazon rainfall simulated by many climate models is much lower than what is actually observed, adding uncertainty to future projections in this region. Past research suggests that the limited model skill is due to insufficient representations of remote ocean influence, evaporation and transpiration by the forest, atmospheric radiation and convection, and other processes. However, the exact mechanisms have not been well understood.

In this study, scientists focused on CESM1 but analyzed many simulations with different configurations to identify the processes underlying the underestimation of rainfall. They found that the complex topography and coastlines surrounding the Amazon are important surface features that induce rising and sinking motions of air and atmospheric waves that distribute the heat and moisture transported from the Atlantic Ocean. They also identified that the way convection responds to winds, water vapor, and temperature has notable influence on Amazon rainfall. When scientists adjusted the way that convective clouds respond to changes in these variables, the rainfall simulation for the Amazon improved. This study and methodology address important considerations in understanding where water vapor goes and how convection is formed in this region with implications around the globe.


Sponsors: The U.S. Department of Energy's (DOE's) Office of Science, Biological and Environmental Research program supported this project as part of the Regional and Global Climate Modeling Program.

User Facilities: This research used computational resources from the National Energy Research Scientific Computing Center (NERSC), a DOE user facility supported by the Office of Science.

Research Area: Climate and Earth Systems Science

Research Team: Koichi Sakaguchi, L. Ruby Leung, Casey Burleyson, Heng Xiao, and Hui Wan, PNNL

Reference: K. Sakaguchi, L.R. Leung, C.D. Burleyson, H. Xiao, and H. Wan, "Role of Troposphere-Convection-Land Coupling in the Southwestern Amazon Precipitation Bias of the Community Earth System Model Version 1 (CESM1)." Journal of Geophysical Research: Atmospheres, 123:8374-8399 (2018). [DOI: 10.1029/2018JD028999].

Page 6 of 740

Science at PNNL

Core Research Areas

User Facilities

Centers & Institutes

Additional Information

Research Highlights Home


Print this page (?)

YouTube Facebook Flickr TwitThis LinkedIn