Modeling Carbon Edge Effects Detected by Remote Sensing Across the Pantropics

• Activity Number: 424
• Type: Contributed
• Date/Time: Tuesday, August 11, 2015 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics and the Environment
• Author(s): Ivan Ramler* and Rebecca Chaplin-Kramer and Richard Sharp and Nick Haddad
• Companies: St. Lawrence University and Natural Capital Project and Natural Capital Project and North Carolina State University
• Keywords: Applied Statistics ; Ecology ; Environment ; Nonlinear regression

Abstract

It has previously been discovered that the proximity to an edge of a forest has a significant effect on the carbon density in trees. However, many of these studies have been limited to determining the degradation of forests due to proximity to forest edge (called an edge effect) for small localized regions where edge effects are typically only observed to be less than 1,000 meters. This study uses remote sensing to model the relationship between carbon storage and distance to forest edge for the pantropical regions of Africa, Asia, and the Americas. Asymptotic growth model (also known as von Bertalanaffy models) are used to estimate edge effects in 100x100km grids throughout the region. Edge effects are consistently found to be present past 1,000 meters, can at times extend to over 5km, and have upwards of 50% reduction in carbon storage between the edge and core of the forest. These edge effects are also linked to biophysical and human factors that influence the degradation of forests.