Mid-summer snow-free albedo across the Arctic tundra was mostly stable or increased over the past two decades

Arctic vegetation changes, such as increasing shrub-cover, are expected to accelerate climate warming through increased absorption of incoming radiation and corresponding decrease in summer shortwave albedo. Here we analyze mid-summer shortwave land-surface albedo and its change across the pan-Arctic region based on MODerate resolution Imaging Spectroradiometer satellite observations over the past two decades (2000-2021). In contrast to expectations, we show that terrestrial mid-summer shortwave albedo has not significantly changed in 82% of the pan-Arctic region, while 14% show an increase and 4% a decrease. The total median significant change was 0.014 over the past 22 years. By analyzing the visible and near-/shortwave-infrared range separately, we demonstrate that the slight increase arises from an albedo increase in the near-/shortwave infrared domain while being partly compensated by a decrease in visible albedo. A similar response was found across different tundra vegetation types. We argue that this increase in reflectance is typical with increasing biomass as a result of increased multiple reflection in the canopy. However, CMIP6 global land surface model albedo predictions showed the opposite sign and different spatial patterns of snow-free summer albedo change compared to satellite-derived results. We suggest that a more sophisticated vegetation parametrization might reduce this discrepancy, and provide albedo estimates per vegetation type.

Automated summary

An analysis of mid-summer shortwave land-surface albedo in the pan-Arctic region over the past two decades shows that the majority of the region has not experienced significant changes, with some areas showing increases or decreases. The slight increase in albedo is attributed to an increase in the near-/shortwave infrared domain, partially offset by a decrease in visible albedo. The findings suggest that a more sophisticated vegetation parameterization method is needed to improve the accuracy of albedo predictions.