Soil Organic Carbon (SOC) is a fundamental indicator of soil health and a key component of global carbon cycling, particularly in arid and semi-arid regions where it is highly vulnerable to mineralization to carbon dioxide (CO2). Hyperspectral imaging (HSI) offers a non-destructive approach for SOC prediction by capturing detailed spectral signatures across the VNIR and SWIR regions. In this study, we developed a comprehensive modeling framework combining laboratory-acquired HSI data with fractional-order derivatives (FODs), multidimensional spectral indices (2D/3D), and the auxiliary use of soil pH, an often overlooked, easily measured, but moderately correlated variable that significantly enhanced model performance. Using 103 dried and sieved soil samples, the optimal feature set that included FOD-based 3D indices and pH achieved the highest test performance (R-2 > 0.85), confirming the synergistic relation between advanced spectral transformations and physicochemical soil properties. To assess the generalizability of our findings, we simulated Maxar-like multispectral bands from the hyperspectral data. Despite the reduced spectral resolution, the simulated Maxar bands achieved an R-2 of approximately 0.65, based on simulations conducted under idealized laboratory conditions with a limited sample size. Notably, Partial Least Squares Regression (PLSR) performed best across our hyperspectral models, in line with its strengths for high-dimensional, collinear spectra in small-sample settings. This study provides a best-case benchmark for translating laboratory spectroscopy to satellite-like bands and indicates potential for scalable SOC mapping pending validation with real imagery and ancillary layers.