Monthly microclimate models in a managed boreal forest landscape

The majority of microclimate studies have been done in topographically complex landscapes to quantify and predict how near-ground temperatures vary as a function of terrain properties. However, in forests understory temperatures can be strongly influenced also by vegetation. We quantified the relative influence of vegetation features and physiography (topography and moisture-related variables) on understory temperatures in managed boreal forests in central Sweden. We used a multivariate regression approach to relate near-ground temperature of 203 loggers over the snow-free seasons in an area of ∼16,000 km2 to remotely sensed and on-site measured variables of forest structure and physiography. We produced climate grids of monthly minimum and maximum temperatures at 25 m resolution by using only remotely sensed and mapped predictors. The quality and predictions of the models containing only remotely sensed predictors (MAP models) were compared with the models containing also on-site measu...