Retrieval of Multiple Variables from Hyperspectral Data: A PRISMA-Aligned Systematic Review of Classical Physics-Based Machine Learning and Hybrid Algorithms in Vegetation and Raw Materials Application Domains

Abstract

Highlights: What are the main findings? Machine learning and deep learning approaches dominate vegetation applications, while classical physics-based methods remain prevalent in raw materials, with hybrid models achieving the highest overall performance across domains. Across vegetation, mineral and methane mapping, insufficient standardization of accuracy metrics and incomplete methodological reporting represent the main sources of bias rather than algorithmic limitations. What are the implications of the main findings? Hybrid physics-informed machine learning approaches emerge as the most promising pathway to ensure robustness, transferability and operational readiness for upcoming hyperspectral missions. The results provide evidence-based guidance for algorithm selection and development aligned with user requirements of current and future hyperspectral missions, such as PRISMA, CHIME, SBG and IRIDE. Hyperspectral (HSI) remote sensing has emerged as a transformative technology for Earth Observation, enabling detailed assessments across different domains. The current PRISMA-aligned systematic review aims to compare classical physics-based algorithms with emerging machine learning (ML), deep learning (DL) and hybrid approaches across two relevant application domains (vegetation and raw materials), analyzing over 350 peer-reviewed studies (194 after the screening) sourced from Scopus and Web of Science and accessed in July 2025. Specific domain-related studies have been considered, excluding duplicates and studies not strictly related to HSI. Risk of bias was assessed qualitatively based on different criteria. The efficiency of the techniques was analyzed by comparing the accuracy metrics reported in the studies. The heterogeneity of the evaluation metrics used across the different categories of the studies and the underrepresentation of some application domains is the final baseline of the work. The results were synthesized, grouping by application domains and algorithm category: ML and DL models dominate vegetation applications, and physics-based methods remain prevalent in raw materials. Hybrid models achieve the highest performances across all domains. This review highlights the importance of the hyperspectral operational requirements identified for upcoming missions (CHIME, SBG and IRIDE) and points out the opportunity for algorithm development.