Abstract | Keywords | Export | Availability | Bookmark

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Afforestation is a proposed tool in the effort to mitigate climate change. However along with sequestration of CO2, the addition of tree cover also leads to biogeophysical forcing on climate through surface changes to albedo, roughness, and latent heat partitioning. As climate warms, we hypothesise that sensitivities of temperature to the change in tree cover would be reduced in winter in historically snow-covered areas, as a decrease in snow cover would lead to less of a change in albedo between forested and unforested land. By conducting a factorial sensitivity analysis using the CMIP6 ScenarioMIP and LUMIP experiments, we found weak results that are consistent with this hypothesis. First, comparing the final 30 years of the historical scenario (1985-2014) and of the future scenarios (2070-2099) the sensitivities were less negative for North America and positive rather than negative for Europe. Next, sensitivities for each continent by linear regression were scaled against the GMT anomaly for 30-year periods overlapping by 5 years spanning the historical (1850-2014) and future (2015-2099) periods. This analysis did not result in a strong trend for the multimodel mean. For North America and Europe there were weak trends following the expected results, from negative sensitivity towards zero sensitivity at higher GMT anomaly. Minimal sensitivity was found for South America and Africa for both the historical and future periods. Asia showed some sensitivity but had mixed results, likely due to conflicting forcings mechanisms due to tree cover change in different climate regions of the continent. To understand how snow cover is related to the changes in sensitivity, the analyses were also repeated using only snow-covered cells, leading to stronger negative sensitivities for North America, Europe and Asia in the historical period and for North America and Asia in the future period. Finally, the sensitivities were scaled against the snow cover per continent, whereby the large spread between models resulted in a limited sensitivity to snow cover based on the multimodel mean. The high level of variability between models presented a challenge for this research and the resulting uncertainty makes is difficult to draw strong conclusions from the results.