Pemodelan Harga Tanah Perkotaan Menggunakan Metode Geostatistika (Daerah Studi: Kota Bandung)

As one of resource factors, land plays a strategic role in urban development. Land price information is needed in urban land management. Spatial distribution of land price can be obtained through spatial modelling. This research aims to examine the use of geostatistical approach in modelling urban land price, with Bandung municipality as the study area. We used ordinary kriging method to predict land price. The semivariogram models used in this research were spherical and exponential models, developed in isotropic and anisotropic approaches. We used 485 samples of market land price data in 2007-2008. Research results showed that spherical semivariogram models gave better accuracy than exponential models. Prediction errors were affected by the amount and distribution of sample data. The spatial distribution of land price of Kota Bandung showed the highest land price occurred in Bandung downtown area, that is, around Jalan Asia Afrika, Jalan Naripan, Jalan ABC, and Jalan Braga. The land price decreased with the increasement of the distance from the downtown area and attained the lowest value in urban fringe areas. The rate of land price increasement did not similar to all directions in Bandung area