I just watched a presentation Elsi Sertel from a university in Turkey showing how easy it is to introduce true land cover data into a climate model. Her study area was around the Black Sea near Istanbul, and used LANDSAT imagery along with a pixel by pixel truthing technique to determine the type of land cover, sea, forest, urban, etc and apply it to use in a GCM.
Her premise was that current GCM’s use land surface info that isn’t fully representative, out of date, and in some cases just plain wrong.
Her study showed a technique that allowed for a significant amount of automation to the process, to allow improved and current land surface types to be easily integrated into the grid cells of a GCM. Unfortunately, some GCM gridding schemes are too coarse to handle such data.
From what I’ve seen in this conference so far, and I’ve seen presentations now from Europe, Turkey, China, Australia and the USA, it is becoming more clear that land use is a major driver of climate change, and perhaps dwarfs even GHG effects. That’s just a hunch. One study from Australia showed the effects of removing a woody type bush over a large area over the past century, and the results on rainfall and temperature were profound.