Understanding of tempo-spatial pattern and the systimatic bias of the obeserved decadal to multidecadal variability and long-term trends of global land Surface Air Temperature （SAT） is needed for climate change studies and policy-making. This paper summarizes the state and problems of the current studies of global land SAT change, and points out the necessarty and possibility to launch a new study of global and regional SAT dataset and analyzing products. It is obvious from the overview that there exist some problems with the current three global datasets under use in global climate change research, and a major issue would be the inefficient treatments of the urbanization bias in the land SAT series. It is proposed that the Chinese global land SAT dataset developed in the China Meteorological Administration （CMA） be improved and completed, and the urbanization effect on SAT trends of global land stations be evaluated and adjusted. Based on the urban-bias adjusted SAT datasets, global and regional SAT series could be constructed and analyzed to reveal the spatial and temporal patterns of SAT variablity and change. Chinese scientists could paly a more important role in the endouvour facing climate community.