Intensive Residential Land Use Evaluation Based On Open Data

The reasonable layout and efficient utilization of residential land not only affect the development of the city,but also related to the daily life and living environment of resident.With the accelerated process of the urbanization in China,the demand for urban residential land has increased significantly.One of the important problems to be solved urgently is how to use residential land intensively.This thesis takes the residential land of Beijing as the main research object,raised the concept of carrying capacity of residential land based on the functional analysis of residential land.Constructed the evaluation indicator system of the intensive use of residential land.The thesis acquired the location and attributed data of residence district and service facility around in Beijing through the web crawler and web application interface,then identified and extracted the land parcel of residential land using twice space superposition analysis method and established a residential land database.Web questionnaire was applied to determine the service radius of different categories of service facility and by using kernel density function to calculate the service capacity of residential land.Finally,nine classes of back-propagation neural networks(BP-NNs)were constructed corresponding to nine classes of land price level of Beijing.The BP-NNs were trained based on the service carrying capacity of residential land(SCCRL)and residential land bear capacity(RLBC).And based on the results of training,we evaluated the intensive use level of residential land of Beijing.At last,the prospective scheme of potential mining was proposed for residential land and recommendations were given for the future policy formulation of residential land in Beijing.In summary,this thesis gives the following conclusions:(1)It is a convenient,fast,and free method to explore the land-use data through the internet.This new approach effectively solves the problem of difficult to access detailed land data in city,and the massive amounts land use or commerce data from internet enrich the attribute data of residential land we used.These methods give us a great opportunity to dig more deep information exist in the land data.(2)A significant correlation can be found between the density and distance of service facilities around the residential and the population,plot ratio,occupied area,and land price of residential land.The more superior of the location of the residential area,the more resident population it carried,and the more intensive service facilities surround.On the contrary,the residential area far away from the downtown of the city,always means deficient service facilities and less resident population.(3)Currently,it is reasonable that the overall utilization level of residential land in Beijing and the intensive use of residential area occupy the principal part.Unreasonable utilization of land is mainly distributed in the junction boundaries of different administrative region,clear partition is forming in the direction of northeast and southwest that is the northwest direction distribute the larger residential area of excessive use,meanwhile the southeast direction primarily has the residential area of inefficient utilization.Different intensity of residential land can be found in different area of land price level and administrative regions.(4)In the future,the residential land in Beijing is still has a great potential for mining.It is suggested that the main development direction of residential construction of Beijing should be toward to southeast,and further improve the intensity level of residential land through city encryption and urban sprawl for promoting the reasonable use of residential land.