A curve-based radiometric calibration method for wavelength of interest hyperspectral imaging: a case study of QMX-1/CCTF-HI

Wavelength of interest (WOI) hyperspectral imaging technology offers significant advantages in optimizing mission-specific goals and improving cost-effectiveness in hyperspectral Earth observation. However, conducting per-band measurements for periodic on-orbit radiometric calibration across thousands of wavebands over the entire spectral range presents a considerable challenge, especially given the limited resources available on satellite platforms and at ground stations. To overcome this, we propose the Curve-based Hyperspectral Imaging Radiometric Calibration (CHIRON) method. This method generates radiometric calibration coefficient (RCC) curves across the imaging spectrum by leveraging the wavelength-dependent optical properties of the system. Integrated within the vicarious calibration workflow of the Radiometric Calibration Network (RadCalNet), CHIRON is applied to the Compact Continuous Tunable LVF-based Hyperspectral Imager (CCTF-HI) onboard the QMX-1 microsatellite. The CCTF-HI is a hyperspectral imager based on a linear variable filter (LVF), operating across the 400–1000 nm spectral range with a 0.287 nm interval, enabling WOI hyperspectral imaging. We employ cubic polynomial regression models to reconstruct the RCC distribution, achieving optimal fits with all models showing R² ≥ 0.92. We validated the radiometric performance of the CHIRON method using three different WOI imaging modes, each capturing a distinct subset of selectable wavebands from QMX–1/CCTF–HI. Reflectance measurements from these modes were compared with RadCalNet products, ground-based spectroscopy, and near-coincident imagery from an independent reference sensor. All absolute errors remained within 0.05 reflectance units, demonstrating that the RCC curve preserves radiometric fidelity even without band–specific calibration. By eliminating exhaustive per–band radiometric characterization, CHIRON reduces calibration workload and computational demand, enabling precise yet resource–efficient WOI hyperspectral imaging – particularly valuable for microsatellite with constrained onboard processing.