Spatial Econometric and Environmental Land Valuation (SEELV): modeling interactional fluxes and net land value in urban zones

Urban land values are shaped by economic intensity, environmental constraints, and infrastructure connectivity. Traditional valuation frameworks often assess these factors in isolation, which limits the comprehension of spatial dynamics and diminishes their utility for planning. This study presents the Spatial Econometric and Environmental Land Valuation (SEELV) framework, which disaggregates urban valuation into source, sink, and structural components, incorporating spatial interactions and model uncertainty. Applied to 129 urban zones in Kuwait, the model produced continuous fields for land value gradients, flux, divergence, and development indicators. These were evaluated using cross-validation, Monte Carlo uncertainty analysis (n = 500 perturbations), and sensitivity assessments. The threshold-parameterized model achieved a root mean square error (RMSE) of 4,505 USD/m2 and a mean absolute error (MAE) of 2,596 USD/m2, surpassing spatial lag, spatial error, and geographically weighted regression (GWR) benchmarks by 4–10% in extreme-condition zones (n = 44). Sensitivity analysis identified nighttime light and texture entropy as predominant drivers (influence scores: 34 and 30 × 10−4, respectively), whereas re-parameterization of environmental sinks increased the sensitivity of drainage density, slope, green deficit, and land surface temperature by 675–1,100%. The SEELV offers a spatially explicit, uncertainty-aware decision-support tool for urban planning and land regulation.