An ecosystem model for tropical forest disturbance and selective logging

[1] A new three-dimensional version of the Carnegie-Ames-Stanford Approach (CASA) ecosystem model (CASA-3D) was developed to simulate regional carbon cycling in tropical forest ecosystems after disturbances such as logging. CASA-3D has the following new features: (1) an alternative approach for calculating absorbed photosynthetically active radiation (APAR) using new high-resolution satellite images of forest canopy gap fraction; (2) a pulse disturbance module to modify aboveground carbon pools following forest disturbance; (3) a regrowth module that simulates changes in community composition by considering gap phase regeneration; and (4) a radiative transfer module to simulate the dynamic three-dimensional light environment above the canopy and within gaps after forest disturbance. The model was calibrated with and tested against field observations from experimental logging plots in the Large-scale Biosphere Atmosphere Experiment in Amazonia (LBA) project. The sensitivity of key model parameters was evaluated using Monte Carlo simulations, and the uncertainties in simulated NPP and respiration associated with model parameters and meteorological variables were assessed. We found that selective logging causes changes in forest architecture and composition that result in a cascading set of impacts on the carbon cycling of rainforest ecosystems. Our model sensitivity and uncertainty analyses also highlight the paramount importance of measuring changes in canopy gap fraction from satellite data, as well as canopy light-use efficiency from ecophysiological measurements, to understand the role of forest disturbance on landscape and regional carbon cycling in tropical forests. In sum, our study suggests that CASA-3D may be suitable for regional-scale applications to assess the large-scale effects of selective logging, to provide guidance for forest management, and to understand the role of forest disturbance in regional and global climate studies.