Persistent spread in seasonal albedo change radiative forcings linked to forest cover changes at northern latitudes

Abstract:

Large-scale land use and land cover change (LULCC) can significantly affect regional climates from changes in surface biogeophysics, and a substantial part of historical LULCC from forest to crop or pasture occurred in the mid- and high-latitudes of North America and Eurasia where the snow-masking effect of forests often leads to a negative radiative forcing from albedo changes linked to deforestation. Results from several recent historical LULCC modeling studies, however, reveal an order of magnitude spread in climate forcing from the snow-masking effect by forests. This is likely because, in months with snow cover, the interactions between vegetation and snow significantly complicate the relationship between the change in forest cover fraction and albedo, thus accurate characterizations of land surface-albedo dynamics are essential given the importance of albedo feedbacks when ground or canopy surfaces are covered in snow

Here, we evaluate snow masking parameterization schemes of seven prominent climate models in greater detail in order to pinpoint major sources of the persistent variability in albedo predictions across models. Using a comprehensive dataset of forest structure, meteorology, and daily MODIS albedo observations spanning three winter-spring seasons in three regions of boreal Norway, we estimate radiative forcings connected to canopy snow masking and compare it to the observed forcings. We develop a physically-based regression model and compare its performance to existing modeling schemes, concluding with a discussion on the utility of purely empirical parameterizations relative to those rooted in radiative transfer theory and/or process-based modeling.