Urban expansion and urbanization have been continuing to grow rapidly, especially in Asian megacities. Greater Jakarta (Jabodetabek) has emerged as the world’s second largest urban area, with a population of 28 million in 2010, where urban expansion has a significant impact on the local as well as the global environment. Efforts to control urban expansion must start from a clear understanding of its various driving forces at a local, regional, and global level. Studies of the interdependencies between these driving forces in the local spatial relationships in emerging Asian megacities remain limited. This study explores the driving forces of urban expansion in Jabodetabek by considering local spatial dependency and analyzes the spatial characteristics of this urbanized area as well as identifies spatial variations in the relationship between urban expansion and its driving forces by using Geographically Weighted Regression. The presented findings show that the driving forces affecting urban expansion in the Jabodetabek region vary spatially. Owing to the influence of the global and regional economies on Jabodetabek, we find that the demographic, infrastructural, and natural elements driving forces significantly affect urban expansion in this region according to location. Outside the core of this megacity, urban expansion in most areas is significantly affected by local demographic as well as infrastructural driving forces. Jakarta city, as the core of the Jabodetabek megacity, is becoming independent of these local driving forces, however, since it is now more characterized as a global city and thus tending to have more linkages with the world market.

Although a number of studies have developed fast geographically weighted regression (GWR) algorithms for large samples, none of them has achieved linear-time estimation, which is considered a requisite for big data analysis in machine learning, geostatistics, and related domains. Against this backdrop, this study proposes a scalable GWR (ScaGWR) for large data sets. The key improvement is the calibration of the model through a precompression of the matrices and vectors whose size depends on the sample size, prior to the leave-one-out cross-validation, which is the heaviest computational step in conventional GWR. This precompression allows us to run the proposed GWR extension so that its computation time increases linearly with the sample size. With this improvement, the ScaGWR can be calibrated with 1 million observations without parallelization. Moreover, the ScaGWR estimator can be regarded as an empirical Bayesian estimator that is more stable than the conventional GWR estimator. We compare the ScaGWR with the conventional GWR in terms of estimation accuracy and computational efficiency using a Monte Carlo simulation. Then, we apply these methods to a U.S. income analysis. The code for ScaGWR is available in the R package scgwr. The code is embedded into C++ code and implemented in another R package, GWmodel.