Abstract: Built-up land sprawl is the most drastic and visual manifestation of urban land-use change. Understanding spatiotemporal change of built-up land sprawl and its driving forces can reveal the inherent laws of urban sprawl, and thus scientifically regulate urban land use. In this study, we used multisource data (e.g., 30 m resolution Landsat images in 1995, 2000, 2005, 2010, and 2015) and socioeconomic data to analyze spatiotemporal changes of built-up land sprawl in Yangzhou City. Further, we employed the boosted regression tree method to quantify the effects of geographic, social and economic factors on such sprawl. Results showed that the built-up land area gradually increased from 725.4 km² in 1995 to 1076.5 km² in 2015. With the increasing urbanization, the spatial form of built-up land tended to be compact and regular, and the gravity coordinates also migrated accordingly, showing local attraction and linkage of municipal districts (e.g., Hanjiang and Guangling). Among geographic, social, and economic factors, geographic factors (e.g., altitude, the distance from the county core, the distance from built-up land) had the greatest influence on built-up land sprawl, with the relative influence gradually decreasing from 73.9% in 1995-2000 to 60.2% in 2010-2015. The influence of social and economic factors gradually increased from 5.2% and 20.9% in 1995-2000 to 7.9% and 31.9% in 2010-2015. Geographic factors (e.g., the distance from the county core and the distance from built-up land) generally had negative effect on built-up land sprawl, but the effects of social and economic factors were inconsistent. For example, population density, GDP per capita and fiscal revenue promoted built-up land sprawl, gross population inhibited it, and secondary industry first inhibited built-up land sprawl and then promoted it.

Key words: Takyr Solonetzs, water content, spectral reflectance characteristics, predict.