Abstract

This paper employs land price gradient and spatial auto-regression for predicting land prices in a satellite town of a mega city. The paper presents the case study of the municipality of Savar, one of the four towns intended to absorb migrants and to confine urban sprawl of Dhaka city in Bangladesh. Like other satellite towns of Dhaka city, Savar is also an industrial town. Three city centers, apart from the central business district, were identified for Savar. It was found that the two additional centers were mainly composed of residential uses, influenced by the proximity to a mega city (Dhaka), the export processing zone, and Jahangirnagar University. Spatial auto-regression analysis revealed a series of factors showing significant weight on land prices of Savar: area and topographical elevations, population density, distance from bus terminal, road infrastructure, central business district, schools, hospitals, markets, and hazardous industries. This case study can serve other municipalities wishing to develop a well-structured land valuation method.