Rédiger un PREreview

This study introduces Δr–NBT, a generalized resolution enhancement framework that upgrades all 16 bands of the Himawari-8 Advanced Himawari Imager (AHI) to 0.005° spatial resolution using physically based interpolation and thermodynamic normalization. During daytime, the method exploits the high-resolution red visible band (Band 03) to compute spatial deltas (Δr) for each visible and infrared band. These deltas are interpolated and applied to the red band at high spatial resolution to generate enhanced outputs. At night, when visible data are unavailable, enhancement is achieved through normalized brightness temperature (NBT), defined as a scaled representation of thermal infrared brightness temperature in degrees Celsius. The framework is fully deterministic and requires no cloud masking, machine learning, or parameter tuning, making it suitable for near, real, time operational use. Δr–NBT is evaluated using a major bushfire event in New South Wales, Australia (10 December 2019), and further validated during a severe urban haze episode in Hanoi, Vietnam (1 November 2024). Enhanced imagery reveals fine, scale cloud and aerosol structures that are not resolved in native, resolution products. Quantitative comparisons with conventional bicubic interpolations yield strong agreement (correlation coefficient r > 0.98, root mean square error < 0.013), confirming radiometric accuracy. Composite products, including true, color, shortwave infrared (SWIR), dust, and brightness temperature difference (BTD) indices, demonstrate improved interpretability across diverse atmospheric conditions. Δr–NBT provides a physically consistent, lightweight approach for generating high, definition geostationary satellite imagery and offers a new standard for AHI, based environmental monitoring.