Estimating Net Primary Production of Swedish Forest Landscapes by Combining Mechanistic Modeling and Remote Sensing

The aim of this study was to investigate a combination of
satellite images of leaf area index (LAI) with processbased
vegetation modeling for the accurate assessment
of the carbon balances of Swedish forest ecosystems at
the scale of a landscape. Monthly climatologic data were
used as inputs in a dynamic vegetation model, the Lund
Potsdam Jena-General Ecosystem Simulator. Model
estimates of net primary production (NPP) and the
fraction of absorbed photosynthetic active radiation were
constrained by combining them with satellite-based LAI
images using a general light use efficiency (LUE) model
and the Beer-Lambert law. LAI estimates were compared
with satellite-extrapolated field estimates of LAI, and the
results were generally acceptable. NPP estimates directly
from the dynamic vegetation model and estimates
obtained by combining the model estimates with remote
sensing information were, on average, well simulated but
too homogeneous among vegetation types when compared
with field estimates using forest inventory data.