Now we’re getting weird.

Johannes Feddema of
the University of Kansas and six colleagues from the US National Centre
for Atmospheric Research report in Science journal that they looked at
changes in land use – the growth of cities, clearing of forests for
agriculture, and draining of marshes – and their impact on climate
change in the next 100 years. They confirmed something
environmentalists have predicted for decades – the destruction of the
Amazon forest would make the local climate 2C (4F) warmer because trees
soak up carbon dioxide from the atmosphere, and burning them releases
it. But then the scientists looked at temperate zones and found the
opposite.

Simulations
predicted the conversion of north American and European forests and
grassland to agriculture would cool the region and counteract the
effects of global warming by 25%-50%. This is because ripening corn and
other staples would reflect more sunlight back into space, and release
more moisture into the air, while dark forests would absorb sunlight
and send thermometers soaring. Ken Caldeira and a Carnegie Institution
team backed the finding in Geophysical Research Letters. "We were
hoping to find that growing forests in the US would help slow global
warming. But if we are not careful, growing forests could make global
warming even worse."

If anyone has access to this paper I’d love to see it. So forests reduce warming because they soak up CO2 but crops reflect more sunlight so no forests reduces global warming.

Have they actually compared the two results to each other? Which effect wins out? Absorbing CO2 or reflecting sunlight?