Geophysical surveys, a means of analyzing the Earth and its environments, have traditionally relied on ground-based methodologies. However, up-to-date approaches encompass remote sensing (RS) techniques, employing both spaceborne and airborne platforms. The emergence of Unmanned Aerial Vehicles (UAVs) has notably catalyzed interest in UAV-borne geophysical RS. The objective of this study is to comprehensively review the state-of-the-art UAV-based geophysical methods, encompassing magnetometry, gravimetry, gamma-ray spectrometry/radiometry, electromagnetic (EM) surveys, ground penetrating radar (GPR), traditional UAV RS methods (i.e., photogrammetry and LiDARgrammetry), and integrated approaches. Each method is scrutinized concerning essential aspects such as sensors, platforms, challenges, applications, etc. Drawing upon an extensive systematic review of over 435 scholarly works, our analysis reveals the versatility of these systems, which ranges from geophysical development to applications over various geoscientific domains. Among the UAV platforms, rotary-wing multirotors were the most used (64%), followed by fixed-wing UAVs (27%). Unmanned helicopters and airships comprise the remaining 9%. In terms of sensors and methods, imaging-based methods and magnetometry were the most prevalent, which accounted for 35% and 27% of the research, respectively. Other methods had a more balanced representation (6–11%). From an application perspective, the primary use of UAVs in geoscience included soil mapping (19.6%), landslide/subsidence mapping (17.2%), and near-surface object detection (13.5%). The reviewed studies consistently highlight the advantages of UAV RS in geophysical surveys. UAV geophysical RS effectively balances the benefits of ground-based and traditional RS methods regarding cost, resolution, accuracy, and other factors. Integrating multiple sensors on a single platform and fusion of multi-source data enhance efficiency in geoscientific analysis. However, implementing geophysical methods on UAVs poses challenges, prompting ongoing research and development efforts worldwide to find optimal solutions from both hardware and software perspectives.

Geophysical surveys, a conventional means of analyzing the Earth and its environs, have traditionally relied on ground-based methodologies. However, up-to-date approaches encompass remote sensing (RS) techniques, employing both spaceborne and airborne platforms. The emergence of Unmanned Aerial Vehicles (UAVs) has notably catalyzed interest in UAV-borne geophysical RS. This study comprehensively reviews the state-of-the-art UAV-based geophysical methods, encompassing magnetometry, gravimetry, gamma-ray spectrometry/radiometry, electromagnetic (EM) surveys, ground penetrating radar (GPR), traditional UAV RS methods (i.e., photogrammetry and LiDARgrammetry), and integrated approaches. Each method is scrutinized concerning essential aspects such as sensors, platforms, challenges, applications, etc. Drawing upon an extensive review of over 435 scholarly works, our analysis reveals the versatility of these systems, ranging from geophysical development to applications over various geoscientific domains. The reviewed studies unanimously highlight the advantages of UAV RS in geophysical surveys. UAV geophysical RS effectively balances the benefits of ground-based and traditional RS methods regarding cost, resolution, accuracy, and other factors. Integrating multiple sensors on a single platform and fusion of multi-source data enhance efficiency in geoscientific analysis. However, implementing geophysical methods on UAVs poses challenges, prompting ongoing research and development efforts worldwide to find optimal solutions from both hardware and software perspectives.

This study introduces a deep learning-based super-resolution approach to enhance the spatial resolution of Sentinel-2 imagery using unmanned aerial vehicle (UAV) data. The proposed scale-adaptive, spatial-attentive, and spectral-preserving enhanced super-resolution generative adversarial network (S³-ESRGAN) extends ESRGAN through three innovations: 1) a scale-adaptive mechanism designed for the 5× upscaling factor between images, 2) a spatial attention module emphasizing salient spatial and structural features, and 3) a spectral preservation component maintaining spectral and radiometric consistency. The architecture integrates dense blocks with spatial-attentive enhancement, multiscale feature extraction, and combined channel-spatial attention mechanisms. S³-ESRGAN was trained on the Sen2UAV dataset containing over 428 000 Sentinel-2 and UAV patch pairs from 33 geographically diverse scenes. The composite loss function combines pixel-wise, perceptual, adversarial, spectral, and edge-preservation terms. Experimental results demonstrate that S³-ESRGAN consistently outperforms state-of-the-art methods, producing sharper textures, enhanced edges, and improved spectral fidelity while preserving radiometric integrity. Ablation analysis further validates the contribution of the spectral preservation module to spectral stability. This research contributes to the remote sensing community by enabling the cost-free generation of high-resolution imagery from open-access satellite data.