Hybrid Classification, Nighttime Light, Random Forest, SCATSAT-1, Urban Mapping.
AuthorsAbstractThe atmospheric dust, extreme climate, and cloudy weather conditions during the monsoon season are the major challenges that are facing the urban monitoring in arid and semi-arid regions such as Rajasthan, where optical remote sensing data is often hard to obtain. In view of these limitations, this study suggests a hybrid approach to regional urban mapping in Rajasthan, India, using scatterometer data from the SCATSAT-1 satellite and VIIRS data (NTL). Areas of built-up and non-built up status were obtained using Ku-band backscatter data (σ0) with a spatial resolution of approximately 2 km. A Random Forest (RF) classifier was trained with well distributed data samples from the whole of Rajasthan collected in the month of May 2017. The classification results of SCATSAT-1 radar data were fused with NTL Radiance thresholding to overcome the inherent ambiguity in radar data between rough dry soil and urban structures. The study illustrates how SCATSAT-1 data when combined with NTL data offers a fast and weatherindependent solution for a large-scale urban monitoring and sprawl assessment in arid regions.
•••••••••••••••••••••••••••••••• ejprd.org - Published by Riset Publication Services LLC
EJPRD
Copyright ©2026 by Riset Publication Services LLC