Fadhli Atarita, PhD

Hyperspectral remote sensing (HRS) has been proven to be applicable across a wide range of geoscience applications, including mineral mapping. However, the inherent complexity of geological targets poses significant challenges in designing effective HRS missions. Additionally, the growing diversity of hyperspectral platforms and sensors, both public and commercial, necessitates careful planning to maximize mission success. Simulation studies offer a cost-effective approach to evaluating the capabilities of hyperspectral sensors and platforms, enabling preliminary assessments without extensive data collection.

In this study, we present HYSIMUv2.0 (HYperspectral SIMUlator version 2.0), a versatile forward modelling toolkit designed to help improve the effectiveness and efficiency of HRS missions. HYSIMUv2.0 allows users to generate at-sensor radiance and reflectance images based on simulated or real ground truth maps. Key features of HYSIMUv2.0 include:  a ground truth data cube generator for customizable input parameters; integrated hyperspectral sensor functions and filters; inclusion of atmospheric effects from radiative transfer models (6S and libRadtran); and scalability for arbitrary hyperspectral sensors, platforms, and spectral libraries. The toolkit is scale-independent and also supports parallelization on High-Performance Computing (HPC) clusters, suited for the extensive computational demands of high-resolution hyperspectral datasets.

This study demonstrates HYSIMUv2.0’s capabilities through case studies that evaluate the resolvability of mineral exploration targets using various platform-sensor configurations. With the continuous advancements in HRS technologies, forward modelling toolkits like HYSIMUv2.0 are essential for optimizing mission designs and assessing the feasibility of survey targets. With its flexibility to accommodate customizable parameters and spectral libraries, HYSIMUv2.0 offers a robust solution for both scientific and industrial applications.