Urbanization is the population shift from rural to urban area, and this is a comprehensive process includes the content of population, ecnomic condition, spatial relationship, and also so many other aspects. It is interesting to mention here that only 3% of the earth surface is covered with urban area, but over 50% of the population residents in cities. Urban spatial expansion is the spatial expression of urbanization, so that it becomes the most important indicator to measure the urbanization level. The rapid urban expansion has resulted in many serious problems such as eco- enviornmental problems. Therefore, it is particular important to examine the state and trend of land use/ land cover and measure the urban extnt and expansion.
Remote sensing data is very cost-effective because of its synoptic view, repetitive coverage and real-time data acquisition. By using the RS(Remote Sensing) and GIS (Geographic information system), the frequent acquisition of massive and detailed regional land use and cover change data can be obtained and help the intensive analysis of the regional urban expansion extent, size, time, rate, character and developing trend.
Digital LULC (land use/ land cover) change detection is the process of determining and describing changes in land-cover and land-use properties based on co-registered multi-temporal remote sensing data. The basic premise in using remote sensing data for change detection is that the process can identify change between two or more dates that is uncharacteristic of normal variation. Numerous researchers have addressed the problem of accurately monitoring land-cover and land-use change in a wide variety of environments.
Material and Method Used in Measuring Urban Expansion
The Landsat Program was launched in 1972 by NASA and provides images free for evaluation of land resources and regional analyzes. The Landsat program consists of a series of eight satellites. Landsat 5 (TM sensor) images were widely used for urban expansion study. The Landsat 5 has polar orbit and circular, sun-synchronous, the area of imaging is the 185 km, with scenes of 185km x 185km. The spatial resolution of the TM sensor is 30m, and the in the thermal band 6, has 120m. The Sensor TM has radiometric resolution of 8 bits and spectral resolution with 7 optical bands: three in the visible portion of the electromagnetic spectrum (blue, green, red); three in the portion of infrared (near infrared and middle), and a band in the thermal infrared. On February 11, 2013 launched Landsat 8 now with 11 bands and the same spatial resolution, equal to the Landsat 5 (except for the thermal bands now 100m).
Landsat images observing at different time were obtained for the further multi temporal analysis. The process of acquiring LULC map can be summarised as: Geometric correction and do the registrations for each of the images of the studying period. Mosaic technique should be utilized when the downloaded raw remote sensing data cannot cover the whole study area, and the area outside the study area can be excluded by the spatial subset operation. It is recommended to do the image fusion if there is high spatial resolution image available such as SPOT or Quickbird. In this way to enhance the accuracy of the land use classification result. Different classification Methods can be applied to produce the LULC maps for the study area.
For the extraction of the urban built-up land, the radiation and atmospheric calibration have to be done in order to obtain the reflectance data for the generation of MNDBI (Modified Normalized Difference Barren Index). As we know that the NDVI (Normalized Difference Vegetation Index) are widely used to reflect the vegetation information of the study area. NDBI (Normalized Difference Barren Index) which takes the reflectance difference of mid infrared and near infrared is widely used to interpret the urban and barren land. The MNDBI is the combination of NDVI and NDBI which can exaggerate the urban built-up land for extraction. Using the MNDBI method, the urban built-up lands are extracted with the thresholding algorithum. By overlapping the boundary of the urban built-up land and analysing the expansion track, we can predict the development of urban expansion with the assistant of the other statistic data.
Monitoring the Urban Expansion
This is the MNDBI map of Shanghai City for the year 1990, 2000, and 2010. We can see that the higher value of MNDBI the more the probability a pixel being a urban built –up land.
Figure 1: MNDBI map; (a) in 1990, (b) in 2000, (c) in 2010
Using the thresholding algorithm, optimal threshold was obtained for the extraction of urban built-up land. . From here, we can see that the built-up land was developing outward continuously. Also, the centroid of the urban built-up land can be calculated in the ArcGIS software as shown in the map.
Figure 2: The location of the built-up areas and their centroids
The area of built-up growth, the annual average expansion rate, and the annual average expansion intensity rate can be calculated. We can also analyze the morphology of Shanghai built-up land using some relative indexes such as compactness ratio and fractal dimension. In terms of the spatial differential analysis, we can see the deviation of the urban built-up land development in different directions, and it illustrates the general trend of urban expansion direction. The results of monitoring the urban expansion gives a general idea of the condition of urban built-up land for the urban planners, and the subsequent suitability analysis can give a clue for urban planners that the urban expansion should be under a certain control to cater for the needs of residents and avoid the overpopulation problems.