SAR-based classification of built-up patterns using Modified PRISI index: Surabaya–Sidoarjo, Indonesia

Urban sprawl in metropolitan areas poses major challenges for urban and regional planning, especially in suburban zones that act as transition areas between urban centres and rural regions. The Surabaya–Sidoarjo area, part of the Surabaya Metropolitan Area (SMA), illustrates this phenomenon, where rapid expansion of built-up areas has significantly altered land use patterns. This study develops built-up area classification approach by modifying the Piecewise Radar Impervious Surface Index (PRISI) and integrating it with the Random Forest algorithm to distinguish built-up areas into regular and irregular categories. Sentinel-1 Synthetic Aperture Radar (SAR) imagery was used as the primary data source due to its ability to overcome limitations of optical imagery caused by cloud cover and atmospheric conditions. The results indicate that non-built-up areas remain dominant, covering 16,776 ha (62.2%), followed by irregular built-up areas at 7,267 ha (27.0%) and regular built-up areas at 2,903 ha (10.8%). Regular built-up areas are mainly concentrated in planned residential zones such as Rungkut District, while irregular built-up areas are prevalent in transitional and peri-urban zones, particularly in Waru District, which is characterized by informal settlements and mixed land use. This classification provides a clearer representation of spatial heterogeneity and highlights contrasts between planned urban development.