A green planet versus a desert world: Estimating the maximum effect of vegetation on the land surface climate

We quantify the maximum possible influence of vegetation on the global clim ate by conducting two extreme climate model simulations: in a first simulat ion ('desert world'), values representative of a desert are used for the la nd surface parameters for all non glaciated land regions. At the other extr eme, a second simulation is performed ('green planet') in which values are used which are most beneficial for the biosphere's productivity. Land surfa ce evapotranspiration more than triples in the presence of the 'green plane t', land precipitation doubles (as a second order effect) and near surface temperatures are lower by as much as 8 K in the seasonal mean resulting fro m the increase in latent heat flux. The differences can be understood in te rms of more absorbed radiation at the surface and increased recycling of wa ter. Most of the increase in net surface radiation originates from less the rmal radiative loss and not from increases in solar radiation which would b e expected from the albedo change. To illustrate the differences in climati c character and what it would imply for the vegetation type, we use the Kop pen climate classification. Both cases lead to similar classifications in t he extra tropics and South America indicating that the character of the cli mate is not substantially altered in these regions. Fundamental changes occ ur over Africa, South Asia and Australia, where large regions are classifie d as arid (grassland/desert) climate in the 'desert world' simulation while classified as a forest climate in the 'green planet' simulation as a resul t of the strong influence of maximum vegetation on the climate. This implie s that these regions are especially sensitive to biosphere-atmosphere inter action.