Savannization of a tropical forest frontier: modeling changes in forest structure from anthropogenic fires and climate change
Douglas
C
Morton, University of Maryland, morton@geog.umd.edu
(Presenting)
Ruth
S
DeFries, University of Maryland, rdefries@geog.umd.edu
The magnitude of current and future carbon fluxes from Amazonia depends on the synergistic effects of climate change and anthropogenic land cover change on vegetation structure. Alone or in combination, these forcing mechanisms have the potential to convert forests into drought and fire-adapted grasslands, a process referred to as “savannization”, with dramatic consequences for carbon storage. The main objective of this research is to evaluate the sensitivity of forests in the southern Brazilian Amazon to savannization through anthropogenic fires for land clearing and management and climate change with a modified version of Ecosystem Demography (ED), an ecosystem model. Initial work on SavannaED includes the development of an anthropogenic fire ignition and spread submodel. Here, we analyze the spatial patterns of historic fires in cerrado and forest cover using coarse and high-resolution satellite data products to identify specific features such as roads or rivers that permit or prevent fire spread on the landscape. We report typical fire sizes in different vegetation types and preliminary findings on the correlation among common landscape features and extent of burned area. Modeling current and future fire susceptibility depends on this understanding of landscape-level features that drive fire risk. Future work will incorporate these fire spread rules into SavannaED to improve the depiction of disturbance and vegetation competition at the forest/cerrado frontier in the southern Amazon. By focusing on the contribution of anthropogenic fires to changes in forest structure, model scenarios will generate insight into how different land use possibilities alter the long-term response of forests to climate change.