Andrea Sealy

Postdoctoral Fellow
TIIMES - CGD - ASP
BGS

Contact Information:
PO Box 3000, Boulder, CO 80307-3000
Office: ML - 593b
Telephone: 303-497-1344
Email: asealy@ucar.edu
CGD Home Page | SERE Home Page | SERE Research

Project Summary:

The impact of dynamic vegetation on dust and Sahel precipitation

Andrea Sealy (ASP/CGD) and Natalie Mahowald (CGD/TIIMES, currently at Cornell University)

I am working with Natalie Mahowald to examine the interaction of vegetation dynamics with desert dust and Sahel precipitation using the Community Atmosphere Model (CAM). The configuration of the model simulations that we are analyzing differ as to whether the runs include dust radiative forcing and feedback (both shortwave and longwave), the SST forcing (whether observed sea surface temperature/AMIP or interactive SST from the Slab Ocean Model/SOM) and if coupled with the Dynamic Global Vegetation Model (DGVM). Comparison of CAM/DGVM to uncoupled CAM simulations suggests that there is a response of precipitation and dust to dynamic vegetation. The interaction with and response of dust and precipitation to dynamic vegetation also appears to vary with observed SST forcing versus interactive SST.

Preliminary results indicate that for runs forced with SOM SST and coupled with DGVM (both with and without dust radiative forcing and feedback), the June-July-August-September (JJAS) precipitation over the Sahel is on average higher than those runs using the default vegetation (Figure 1, left).

However for the cases forced with observed SST the coupling with DGVM produces more complex results. The JJAS precipitation is lower for the DGVM cases over sub-Saharan Africa, but in the coastal regions closer to the Gulf of Guinea, the JJAS precipitation is higher (Figure 1, right).

In examining the area average over 12°N-18°N, 18°W-20°E the SOM cases with DGVM have higher mean JJAS precipitation than those cases without DGVM, for the AMIP cases the result is the opposite (Table 1). For both AMIP and SOM cases, the amount of dust decreases when CAM is coupled with DGVM (not shown).

We are also analyzing other variables such as surface albedo, surface and atmosphere shortwave and longwave radiation, temperature, plant functional type (pft) and leaf area index (LAI).