Urban Sprawl Dynamic Monitoring Using Landsat TM/ETM～+ And GIS In Jinan City

As accelerating economic growth, rapid urban sprawl and dramatic change of land use structure became an important issue. In order to control the urban rational scale, it is necessary to monitor the urban sprawl in time. With the development of remote sensing (RS) and geographic information system (GIS), RS and GIS techniques are applied more and more in urban land use change detection. Study area lies in Jinan City, Shandong Province. On the basis of collecting and summarizing literatures, we carried out monitoring of land use change by RS and GIS. The RS Data we used is Landsat TM/ETM+ acquired at 1987.5.14,1996.10.11 and 2001.9.1. The study focused on the methods of urban sprawl dynamic detection, the urban expansion spatial pattern and driving forces.On the basis of geometry registration, enhancement and data fusion for TM/ETM+ images, we tried to dectect urban expansion by three methods: post-classification comparison, threshold of Normalized Difference Vegetation Index (NDVI) and principal component transformation. (1)Change detection based on post-classification comparison. It is detecting land use change by contrasting classification results of RS images and cofirming the variational area. This method is not affected by the seasonal temporal of TM/ETM+, however the result of monitoring is exaggerated due to the accumulation of error.(2)Change detection based on threshold of NDVI. Firstly, we need to confirm the area that is not covered by vegetation using the threshold of NDVI,and then extract the urban area boundary. Comparing the urban area of different images, the change areas are determined. This method needed to choose the images that have same spectrum character. And the main issue is how to determine the threshold according to the histogram of NDVI.(3) Change detection based on principal component transformation. That is synthesizing the multi-temporal imges to confirm the area of urban sprawl. Two methods are used: multi-band principal component transformation and subtracted principal component transformation. This method is good when the land use change is obvious enough to detect. It needed to analyse the significance of each PC and confirm which PC is used to monitor urban expansion.Using the urban sprawl information from 1987 to 2001 by RS and the urban map of Jinan City in 1977, we studied the urban expansion spatial pattern and its driving forces by means of GIS technique. The results show that:(1) In the period during 1996~2001, the annual increase area is more than any other period. But the general ratio of urban sprawl tends to decline.Separated by the Yellow River in the north, the main urban expansion direction are east ,east-north, west,south and west-south.(2) By buffer analysis, we found that the urban expansion pattern is concentric circles. The urban sprawl in Jinan is affected by economy development, population increase and traffic condition improvement. And the economy development is the main factor...