Near-surface geophysics: from measurements to reliable models

Abstract

Near-surface geophysics encompasses methods for investigating the upper layers of the Earth’s crust, where geological processes directly affect infrastructure, natural resources, and environmental risks. The Near-surface geophysics Collection brings together studies that demonstrate the rapid evolution of the field, ranging from advances in data acquisition and processing to new inversion and regularization strategies and integrated case studies. In this editorial, the contributions are grouped into three themes: improving the coverage and quality of seismic data; developing inversion and regularization methods for diverse datasets; and integrating data from multiple geophysical methods. Together, these approaches highlight the move toward more reliable and higher-quality subsurface models.

Introduction

Geophysics is a branch of the Earth sciences that investigates the Earth using methods derived from physics¹. Geophysical research focuses on processes occurring within the lithosphere, atmosphere, and hydrosphere. The Near Surface Geophysics Collection was designed to provide an overview of recent advances in methods, approaches, and applications for studying the shallow subsurface, typically the upper hundreds of meters of the Earth’s crust. Its scope encompasses seismology, magnetism, gravimetry, geodesy, and a wide range of remote-sensing and satellite-based techniques ^2–4.

Applications of near-surface geophysics span both socioeconomic and scientific domains, including hydrocarbon exploration, engineering and geotechnical investigations, and research in hydrogeology, archaeology, and environmental sciences. The articles in this collection can be grouped into three thematic categories: (i) new strategies for data processing and representation aimed at extracting reliable information from limited or noisy datasets, (ii) development of inversion and regularization methods, and (iii) integration of data from different near-surface geophysical techniques.

New strategies for data processing and representation

Field seismic surveying has accelerated in recent years with the adoption of mobile systems; however, these devices have limitations, such as reduced ground coupling and limited penetration depth ^5–7. Hanafy proposes a method to virtually extend the land streamer’s data-acquisition aperture without additional hardware or field deployment, thereby extrapolating the aperture without adding equipment⁸. The method has been tested on synthetic models, and the results indicate high accuracy in reproducing far-offsets, potentially enabling deeper and more reliable imaging in engineering and environmental applications.

When the signal characteristic is challenging, stable detection of temporal attributes is crucial. Zhang et al. focus on automatic arrival time detection in low SNR (signal-to-noise ratio) microseismic data⁹. The authors propose a multi-window energy quotient method that is improved by differential transformation. In practice, these solutions matter because initial decisions in the processing chain can significantly influence overall interpretive uncertainty.

Liu et al. present examples of innovation in radar signal analysis, a combination of a short-time canonical linear transformation with joint analysis in the time, space, and frequency domains to get a clearer view of subsurface information in GPR (ground penetrating radar) data¹⁰. This is especially relevant given the rising popularity of time–frequency methods, which focus on filtering noise and improving the separation of overlapping wave phenomena.

Surface studies often occur in extreme environments where “classical velocity” alone cannot fully characterize the medium. Eleslambouly et al. examine sabkhas (saline formations), both inland and coastal, using V_P (P-wave velocity), V_S (S-wave velocity), V_P/V_S ratio, seismic anisotropy, and Q_P⁻¹ attenuation analysis¹¹. This set of parameters are closely aligned with the physics of a porous, brine-saturated medium. This approach strengthens the connection among geophysics, geotechnics, and hydrogeology and enables us to discuss the medium in terms of mechanics and flow rather than solely in terms of layer geometry.

Inversion and regularization of the data

Another issue in analysing near-surface geophysical data is the use of inversion and regularization techniques, which address the well-known geophysical paradox of having many models but little data. Laby et al. [missed citation link: https://www.nature.com/articles/s41598-025-30141-y] show that regularized ensemble Kalman inversion effectively balances robustness and efficiency in modeling gravity data for mineral and ore deposit identification. By using Tikhonov regularization, this method stabilizes the inversion process and improves the Kalman gain’s conditioning, preventing unstable solutions amid noise and data ambiguity.

Conversely, Zhou et al. concentrate on transient electromagnetic method (TEM) data inversion by sparse regularization, tackling the common issue of how to identify detailed boundaries and small targets in the context where smooth regularization blurs contrasts¹². This is an important perspective in the ongoing debate over whether model stability or the identification of local anomalies is more important, particularly in high-resolution prospecting.

The subject of computational methods is addressed by the research of Mahanta and Tiwari, who demonstrate the possibility of using AI for analysing spatial relationships in geophysical fields and grid models¹³. This research integrated computational modelling and geomorphological imagery to develop pre-trained CNN-based prediction tools that estimate the probability of fairy circle occurrence from satellite imagery. After rigorous training and testing, the models achieved up to 98% accuracy in identifying areas with fairy circles.

Data integration in the near-surface geophysics

Methods evolve most rapidly when they align with geological needs and are validated in the field. Qinhong et al. employed a two-dimensional discrete wavelet transform to separate field sources from high-precision grid data, and discuss the separation of sources in the field and the nature of deep and shallow tectonics¹⁴. This research underscores that, in near-surface geophysics, the ‘shallow-deep’ boundary is often not sharp, and extracting shallow information sometimes requires separating contributions from different scales and depths.

Aldeep et al. show that combining methods allows us to simultaneously capture structural elements (magnetics) and the medium’s electrical properties, thus providing answers that are important not only scientifically but also for water resource management and planning in arid environments¹⁵. Abdelrahman and Alarifi demonstrate the use of Multichannel Analysis of Surface Waves (MASW) for geological characterization in terms of sustainable urban development¹⁶.

Summary

The set of analysed articles published in this Collection demonstrates significant progress in near-surface geophysics research in recent years. Application of advanced methods for analysing data, e.g., interferometry and time–frequency analysis, or managing uncertainty in inversion (ensemble, geostatistics, sparsity regularization, and careful use of deep learning models) improves the quality and reliability of solutions. Moreover, near-surface geophysics increasingly operates under time, social and environmental constraints (risk, safety, costs, environment), which reinforce the need for transparent computational workflows, reporting of key processing parameters, and, where possible, the sharing of data and code. In this context, work that demonstrates methods that can be implemented operationally, rather than just conceptually, is particularly valuable.

Author contributions

Kamil Maciuk: writing—original draft preparation; writing—review and editing, Arkoprovo Biswas: writing—review and editing, Shashi Kumar: writing—review and editing, Abhishek Kumar Rai: writing—review and editing. Percentage contribution: KM 70%, AB 10%, SK 10%, AKR 10%.

Declarations

Competing interests

The authors declare no conflicts of interest that could influence the views expressed in this editorial. Their opinions are independent and impartial.