Multifractal characterization of urban residential land price in space and time

Abstract

The spatial and temporal distribution of land price plays a key role in urban development and redevelopment processes. Identifying the features of land price distribution (LPD) is essential for improving urban planning and modeling land use changes. The purposes of this study were to determine if LPD can be characterized by multifractal models and to develop multifractal methods for characterizing the properties of LPD at various scales. An analysis was performed for a study site in Wuhan City, central China. Land prices were sampled in the years 2001, 2004, and 2007. The LPD patterns were represented by multifractal spectra estimated using the method of moments and characterized by five quantitative multifractal parameters. The results showed that the dimension spectra calculated from the LPD data in various regions and at different times indeed depict multifractality, the curves of the multifractal spectra are continuous, displaying the same characteristics of asymmetric and convex curves at the same times in different regions, where the common transitional trend was from shorter toward the left but much longer toward the right in 2001, comparatively symmetric with a slight right deviation in 2004, shorter toward the right but much longer toward the left in 2007, which implying continuous multifractality observed for LPD, and this trend indicates that the singularity of land prices in the different areas keeps step with urban development. In addition, the horizontal characteristics of the curves also differed in different development stages in the city. These results also demonstrated that we may characterize the spatial and temporal differences of different LPD patterns using multifractal methods, which may thus be utilized as a quantitative measure in understanding how land price affects changes in urban land use.