(INVITED) Research Issues raised by looking at the dependence of Tibetan climate on albedo from perspective of a Global Climate Model

The Tibetan plateau in simulations from global climate models has commonly been identified as a region of large disagreement with observations.

We have examined why in a state of the art climate model (CCSM3). The modeled climate has too much winter and spring snow and is too cold at the surface. We also notice that it has a much higher albedo than observed from the MODIS satellite. We have done a sensitivity simulation testing the effects of inserting the observed albedo. With the observed albedo, the surface temperature is in much better agreement with observations. The increased heating from the lower albedo lowers the geopotential height over Tibet and to the NE by about 50 m. An examination of the diurnal boundary layer at the Anduo station indicates reasonable agreement with GAME –Tibet field observations but a strong sensitivity to antecedent and ongoing precipitation conditions – daytime boundary layer height for dry days can be 4 km above the 5 km surface but be 0 for wet conditions. The issue of the large difference between modeled and observed albedo suggests the importance of various complex terrain processes that cannot yet be included in the climate model. An approach under development for treating the subgrid radiative and boundary layer variability is described