Evaluation of the effect of selective logging on the energy-water and carbon exchange processes

This paper discusses the effect of selective logging on the energy, water, and carbon exchange of tropical forest. We apply multi-objective sensitivity analysis and parameter estimation procedures (MOGSA-UA and MOSCEM-UA) developed at the University of Arizona, USA, to the Simple Biosphere Model 2 (SiB2) at a single site in the Amazon Basin (specifically, the Santarém km 83 - LBA site) under two different conditions, i.e. before and after selective logging of the natural forest. It is assumed that logging did not change soil parameters and the results confirm our working hypothesis that the limited changes in the vegetation cover also do not greatly affect the preferred model parameter values in these two cases. However, the results do show that parameter identification procedures are able to retrieve meaningful values for the parameters and do yield an improvement of between 30 and 70% in the root mean square error when compared to using the default parameter values in SiB2.