Use of Remote Sensing to Model Land Use Effects on Carbon Flux in Forests of the Pacific Northwest, USA

Abstract

Reducing the uncertainty in the global carbon (C) budget will require better information on regional C budgets. The use is discussed of a simple 'metamodel', in conjunction with satellite data, to quantify C flux from a 12 000-km2 forestland study area in Oregon, USA, consisting of a mixture of natural and managed stand types and ages in the western hemlock (Tsuga heterophylla) vegetation zone. The model tracks C storage in living, detrital and forest products pools. Between 1972 and 1991, annual total C flux from this study area to the atmosphere was estimated to average 1.13 Mg ha-1. This spatial variability was related to site quality, land use and historical factors. These results are used to illustrate the natural and anthropogenic sources of heterogeneity that can influence C budgets at the regional scale and to demonstrate how remotely sensed data can be used to help quantify this heterogeneity.