| Urbanization has profound influence on today‘s world. The urbanization in China has been into a period of accelerated development. Rapid urbanization is not only in the proportion of urban population increase, but also brings the changes of urban landscape pattern, industrial structure, urban land use and employment. With the further development of urbanization, the urban ecology in China has faced with severe challenges. Therefore, significant attentions has given to urban landscape and its change. Urban landscape pattern is the reflectance of human being and land environment system interactions on geographical space. Spatio-temporal changes of urban landscape will reflect the process of urbanization. Studying the spatio-temporal changes, evolutions and the formation mechanism of urban landscape pattern have an important theoretical and real significance for achieving sustainable development of the city.In this study, the Urumqi city was chosen as study area, firstly we combined using ALOS/AVNIR-2、ALOS/PALSAR and ASTER GDEM by decision tree method, extracted the main city landscape information and the accuracy of classification result has been evaluated. Because the lack of collected data, it is difficult to analyses the spatio-temporal changes of urban landscape. Therefore, using Remote Sensing, Geographical Information System and the landscape ecological theory, landscape pattern dynamic changes and spatial conversion features in Urumqi city during last 20 years were investigated using three Landsat TM/ETM images of Urumqi city, acquired in 1990, 1999, 2010 and DEM image respectively. By using the U.S. Terrain map Earth Imaging LLC developed remote sensing image processing software CPANCROMA software platform(www.terrainmap.com), the gap on Landsat7 ETM + in 2010 were successfully repaired. The main results are shown as blow:(1) Integrated using of optical and microwave remote sensing data, could complement the defects of same spectrum reflectance with different land classes on optical data, for example, the rock, bare land and mountain area will misclassified as urban construction area. This method could provide efficient and more accurate information for extracting urban area, the overall classification results was 91.5%.(2) The analysis of urban landscape spatio-temporal dynamic results showed that, during the whole period, landscape pattern of study area has changed a lot. Especially the construction land was increased dramatically and the land use dynamic index was 16.87%, forestland was increasing trend, land use dynamic index was 13.2%, other vegetation land and unused land were decreased continually, decreased 1.23% and 1.75% per year. On the transition of landscape types, there happened some change on spatial distribution of landscape types. Mainly the increase of construction land and the decrease of unused land and other vegetation. The increase of construction land was mainly come from the other vegetation and unused land, 51.32 km2 and 150.85km2 area were changed to construction land separately. Because of the restriction of topographic conditions, city was expanded significantly in "T" axial type, the space of each landscape type conversions was mainly happened in the northwest and northeast region of the city. On the level of landscape types, landscape dominance of construction land was increased, boundary was becoming more complex, and the polymerization degree was increased. Other vegetation landscape dominance index was decreased, fragmentation degree was increased. Landscape shape of water was inerratic, and fragmentation degree was decreased. Unused land’s fragmentation degree, landscape advantage degree and aggregation degree was declined. On landscape level, the whole city landscape fragmentation degree was increased, landscape shape was tends to be more complicated, landscape heterogeneity was enhanced, that showed the city landscapes development trends towards homogenization and diversity.(3) The analysis of urban greenbelt landscape results shows that, main types of green space are public greenbelt and unit affiliated greenbelt. The urban greenbelts are low diversity, there are few public greenbelts and they distributed collectively, as a result they forms numbers of greenbelt blind areas. Each greenbelt landscape types are have high fractal dimension and their shape was irregular, unit affiliated greenbelt and residence greenbelts fragmentation index are quite high, due to the effects of humans, the shapes are various and irregular.(4) The main driving force of urban landscape change was demographic factors, economic factors, industry structure and political factors, etc. The effects of regional terrain factors has also plays a restricting role on landscape dynamic change, and in some extent will allocate the development framework of urban space compositions. |
PDF Full Text Request