Progress in the study of scale issues in land change science
土地变化科学中的尺度问题与解决途径

Scale issues are the focus of land change science. Systematical summary of the progress of the scale issues in land change studying at home and abroad for recent ten years shows that: the scale issues in land change study mainly focused on data processing, pattern and process representing, the influence of driving forces, application of models, environmental effects, land policy and sustainable management. Scale issues arise primarily from the heterogeneity of geographical phenomena, hierarchy of geographical systems, non-linear characteristics of the response and feedback, influence of the disturbance, limitations of subjective awareness and so on. The scale issues have puzzled the development of land change science, Land Use and Land Cover change (LUCC), Global Land Project (GLP), Millennium Ecosystem Assessment (MA) and other major international projects, which all have taken the scale issues as a prior concern in its research agenda. It is urgent that multi-case synthesis, multi-level data processing, understanding of the relationships between global and local scales and the application of landscape ecology research paradigm should deal with scaling. The general approach of scale study is scale selection - scale analysis - scale integration; when performing the scale selection, we should consider the questions which would be solved and the data which can be collected, then choose the appropriate scale; when performing the scale analysis, it is required that analysis should be carried out on larger and smaller scales at the same time, to find a significant change in dynamics, to prevent the omission or exaggeration of information. Scaling is the bond to recognize the relationship between global and local scales, which can be divided into scaling-up and scaling-down, when performing the scaling, the method is the key problem, whose objective is to find the "connectivity" between scales. Models can help to deeply understand the land use system dynamics, thus developing the nested model is the key problem in scaling study nowadays, for example, to nest the CLUE-S with ABM/LC and so on.