A Hyperspectral Fast Radiative Transfer Model with Controllable Accuracy (HFRTM-CA) and its application in Geosynchronous Interferometric Infrared Sounder (GIIRS)
摘要
This study addresses the increasing demand for hyperspectral radiative transfer models, driven by advancements in hyperspectral instruments and the growing accuracy requirements for applications such as satellite payloads, weather forecasting, and climate modeling. The Hyperspectral Fast Radiative Transfer Model with Controllable Accuracy (HFRTM-CA) model was developed to tackle the challenges of high computational complexity, slow processing speed, and large storage requirements typically associated with hyperspectral radiative transfer and atmospheric transmittance calculations. Using the Geosynchronous Interferometric Infrared Sounder (GIIRS) instrument for validation, the model demonstrated the ability to balance computational efficiency with high accuracy. HFRTM-CA utilizes a comprehensive training dataset to ensure thorough coverage, improving data filling and facilitating future applications. By independently setting accuracy thresholds for atmospheric transmittance and monochromatic radiance, the model optimizes the trade-off between speed, storage, and precision. The use of Principal Component Analysis (PCA) further reduces data storage needs. Test results reveal that HFRTM-CA achieves high-precision simulations, with radiance errors within 10− 3 W/m2/Sr/cm− 1 and channel brightness temperature deviations under 0.2 K. Compared to other models, it provides more accurate results and is easy to set up and apply across various instruments and atmospheric conditions, making it a versatile tool for scientific and operational use.