Impact of Vegetation on the Global Monsoons

The impacts of vegetation on the global monsoon systems, including over North America, South America, West Africa, China, India, and northern Australia, are examined using the fully coupled global climate model, NCAR's Community Climate System Model Version 3.5 (CCSM3.5). The model is run as finite volume with a 1.9x2.5 degree horizontal resolution. Dynamic vegetation is represented by the Community Land Model (CLM) for sub-annual processes and Lund-Potsdam-Jena (LPJ) model for annual processes.

First, the ability of CCSM3.5 with interactive vegetation to simulate these global monsoon systems is evaluated. In general, the model produces a reasonable seasonal cycle of precipitation in these regions, although the monsoon rainfall is typically too great and the rain belt penetrates too far inland. Next, the simulated vegetation distribution is evaluated and found to be reasonable in capturing the large-scale biome patterns that are observed. Over the U.S., the model successfully produces grasslands in the West, deciduous forests in the Northeast, and evergreen forests in the Southeast. Globally, key biases include large holes in the boreal forests and too few trees in the Amazon, while most other forests are too dense.

In order to assess vegetation feedbacks over the global monsoon regions, initial value ensemble experiments are performed in which the vegetation cover fraction over these regions is reduced by 0.2 at the start of each ensemble and the climatic response identified. Generally, a reduction in vegetation over the monsoon regions results in higher temperatures and a complex precipitation response unique to each region. Further analysis over the North American monsoon reveals that a statistical feedback analysis using lagged covariances yields a similar feedback strength to the dynamical, ensemble method. Time permitting, impacts of Rocky Mountain snowpack on the North American monsoon will be explored.