Characterizing urban growth in Vientiane from 2000 to 2019 using time-series optical and SAR-based estimates of urban land

Abstract

While annual urban growth rates are declining in many highly urbanized regions of the world, Southeast Asia (SEA) is experiencing an accelerated urbanization process in small and medium-sized cities (SMCs), to which little attention has been paid. High-precision urban land information is a prerequisite for analyzing urban expansion characteristics for SMCs. However, accurately extracting urban land information for SMCs in SEA has remained challenging, considering the lower urban land density, higher vegetation cover, and complex landscape mosaic. This study systematically analyzed the capacities of a set of statistical indicators of time-series optical and synthetic aperture radar (SAR) images in impervious surface extraction. Then, an automatic urban land mapping method was developed by integrating time-series optical and SAR statistical indicators to map urban land area at a finer spatial resolution. The method was applied to Vientiane, Laos, as a case to explore the characteristics of the rapid urbanization process over 2000–2019 and problems related to sustainable urban development. Our results show the statistical indicators of SWIR1, SWIR2, VV, and VH can distinguish urban land from bare soil and cropland, and statistical indicators of NDVI can separate urban land from seasonal vegetated land. The overall accuracies of our products on Vientiane exceeded 95%, and the kappa coefficient was close to 0.90. The study found that the city was experiencing accelerated urban growth, and the urban land area increased from 25.93 km2 in 2000 to 37.23 km2 in 2010 and 62.12 km2 in 2019. Spatially, the urbanization patterns of Vientiane showed a certain degree of sprawl, especially in the suburbs. Rapid urban sprawl poses a significant threat to the urban environment and sustainable development.