Advancing Urban Sustainability through Remote Sensing and Environmental Valuation

Rapid urbanization has intensified socioeconomic inequality, environmental pressures, and demand for sustainable urban planning. While remote sensing, environmental economics, spatial analysis, and artificial intelligence provide valuable tools for understanding these challenges, their integration across different dimensions of urban sustainability remains limited. Our research addresses these gaps through five interconnected studies in Bengaluru and Kolkata, examining socioeconomic inequality, environmental valuation, neighborhood environmental quality, and human-perceived greenery. The analysis first examines whether hyperspectral remote sensing provides advantages over multispectral imagery for urban socioeconomic mapping. PRISMA hyperspectral and Landsat 8 multispectral imagery were evaluated using supervised machine-learning methods. PRISMA outperformed Landsat 8 in 14 of 16 model–date comparisons, with maximum accuracies of 76.71% in Bengaluru and 74.26% in Kolkata. Building on the socioeconomic mapping analysis, the subsequent studies examine the valuation of urban green spaces from both stated- and revealed-preference perspectives. This valuation is examined by considering the additional percentage of the housing purchase budget that residents are willing to allocate for a residential property located near publicly accessible urban green spaces, using the Contingent Valuation Method (CVM). Based on 602 valid survey responses from Bengaluru and Kolkata, factor analysis, logistic regression, ordered logistic regression, and OLS were used to examine willingness to pay. The mean stated housing-budget premium was 11.87%, and 74.75% of respondents were willing to allocate at least 10% additional housing expenditure. Perceived benefits and policy and institutional support were positively associated with willingness to pay, while lower-income groups reported lower stated willingness to pay. Following the stated-preference analysis, actual housing-market transactions are used to examine whether the value of environmental amenities is reflected in residential property prices. Using 4,665 apartment transactions in Bengaluru from January to August 2024, hedonic pricing and spatial econometric models were employed. Distance to green space was consistently negatively associated with apartment prices, with a 100-meter increase in distance associated with approximately a 2.1–4.5% decline in prices. Blue-space effects were comparatively weaker and less consistent across spatial models. The housing-market analysis is subsequently extended beyond proximity to individual environmental amenities to consider the broader environmental conditions surrounding residential properties. EVI, NDBI, and LST were derived from Landsat imagery and incorporated into spatial hedonic models. EVI remained positive and statistically significant across OLS, SEM, and SAR, whereas NDBI was not statistically significant, and LST was significant only in OLS. Vegetation quality, therefore, showed the most consistent relationship with housing prices among the environmental indicators considered. Finally, the analysis considers environmental quality from a human perspective by examining street-level greenery. The Average Green View Index (AGVI), derived from street-level imagery using Grounding DINO and SAM. Spatial econometric models, XGBoost, and SHAP were used to assess its relationship with housing prices and predictor contributions. AGVI provides complementary information to satellite-derived vegetation measures, and its association with housing prices varies across urban contexts, with the AGVI–IT-zone interaction remaining positive and significant after accounting for spatial dependence. Collectively, these studies bring together socioeconomic mapping, stated and revealed environmental valuation, satellite-derived environmental indicators, and street-level greenery to provide a more comprehensive assessment of urban socioeconomic and environmental conditions. The findings provide evidence relevant to urban planning and environmental management, particularly for socioeconomic monitoring, green space planning, and assessing neighborhood environmental conditions. The research also contributes evidence relevant to Sustainable Development Goal 11, particularly Target 11.7, which emphasizes access to safe, inclusive, and accessible green and public spaces..