Hyperspectral narrowband imaging spectroscopy: A new paradigm for Earth observation

Abstract

This editorial introduces the Special Issue entitled “Hyperspectral Narrowband Imaging Spectroscopy: A New Paradigm for Earth Observation” in the August 2026 issue of Photogrammetric Engineering & Remote Sensing (PE&RS), the flagship journal of the American Society for Photogrammetry and Remote Sensing (ASPRS). This volume represents the fourth dedicated hyperspectral special issue published in PE&RS, following earlier contributions by Thenkabail et al. (2025, 2024a, 2024b), and continues ASPRS’s commitment to advancing cutting‑edge imaging spectroscopy research and its applications across Earth system science. Remote sensing is undergoing one of the most profound transformations in its history. The emergence of hyperspectral narrowband (HNB) imaging spectroscopy data, capable of acquiring hundreds of contiguous, narrow spectral bands, has shifted the discipline from observing Earth in a handful of broad spectral windows to capturing continuous spectral signatures of the Planet. This transition marks a decisive break from the multispectral paradigm that has dominated satellite remote sensing for nearly five decades, driven by the advent of new orbital imaging spectrometers such as EnMAP, PRISMA, and NASA’s EMIT, and by the forthcoming Surface Biology and Geology (SBG) mission ((Pires Silva et al., 2026; Bourriz et al., 2025; Thenkabail et al., 2025; Chabrillat et al., 2024; Aneece et al., 2024; Dave et al., 2024; Thenkabail et al., 2024a; Thenkabail et al., 2024b; Thenkabail, 2024a; Thenkabail, 2024b; Thompson et al., 2022; Kokaly et al., 2022; Aneece & Thenkabail, 2022; Cawse Nicholson et al., 2021; Guanter et al., 2021; Vangi et al., 2021; Thenkabail et al., 2021). These missions (e.g., Table 1) deliver unprecedented spectral fidelity, improved signal to noise ratios, and global coverage capabilities, enabling a new era of quantitative, spectroscopy based Earth observation. Where multispectral broadbands (MBBs) provide only a few discrete measurements along the electromagnetic spectrum, HNB systems deliver rich, diagnostic information that enables scientists to characterize Earth’s surface with unprecedented biochemical, biophysical, and structural detail (Figure 1a, 1b). The implications for environmental monitoring, agriculture, water resources, and mineral exploration are profound. Several overarching themes emerge: • Spectral fidelity matters. The ability to preserve subtle absorption features is essential for mineral mapping, vegetation trait retrieval, and biochemical modeling. • AI and deep learning are indispensable. From destriping to classification, modern analytics must be scalable, label‑efficient, and capable of exploiting the full spectral–spatial richness of HNB data. • Physics‑based and data‑driven approaches must converge. Radiative transfer models such as PROSAIL, enhanced with localized soil parameterizations, remain foundational for biophysical retrievals and model‑based inference. • Dimensionality reduction and feature extraction are critical. Techniques such as L1‑ISOMAP demonstrate that intelligent manifold learning can unlock the structure of fused, high‑dimensional datasets. • Next‑generation architectures must be interactive and multimodal. ICTNet exemplifies the future of hyperspectral classification: hybrid, synergistic, and capable of modeling both local textures and global spectral dependencies.