Research Review · Geothermal Geophysics
Mapping the Mediterranean’s Hidden Heat — Why No Single Method Sees a Geothermal System
Every geothermal prospect in the Mediterranean sets the same trap: the geology that makes it attractive is the geology that makes it ambiguous. A conductive anomaly may be a productive clay cap — or barren saline sediment. A low-velocity zone may be hot fractured rock — or simply a change in lithology. This review sets out why one technique can never resolve that ambiguity, and how seismic, electromagnetic and near-surface methods — each assigned a specific, testable role — together convert a persuasive image into a defensible decision.
The Challenge
Geothermal energy is drawing renewed attention across Europe as a source of low-carbon baseload electricity and dispatchable heat. In the Mediterranean the opportunity is unusually rich: high heat flow, active volcanism, extensional tectonics and fault-controlled hydrothermal circulation produce everything from high-enthalpy volcanic fields to low- and medium-enthalpy sedimentary resources. The catch is that the same complexity that makes these systems attractive makes them hard to characterise.
A productive reservoir depends on the interaction of heat, permeability, fluid chemistry, phase state and structural inheritance — and these vary from metres around a well pad to tens of kilometres across a volcanic or graben system. No borehole samples that range. Geophysics is the only practical way to image the three-dimensional architecture of an active system between sparse wells, but it is intrinsically a multi-physics problem:
- Seismic methods respond to elastic properties, fracture density, attenuation and fluid phase.
- Electromagnetic methods respond to conductivity contrasts tied to hydrothermal alteration, salinity and temperature-dependent fluid connectivity.
- Radar (GPR) responds to dielectric contrasts, at much shallower depths but far higher resolution.
Each property is diagnostic of something — and none is diagnostic of everything. A single dataset can always be re-interpreted. The most reliable geothermal models emerge only when these data types are integrated rather than used in isolation. The figure below shows why: laid out against depth, the methods do not compete for the same ground — they occupy complementary, and only partly overlapping, parts of the system.
Why Integration, Not a Single Method
The physical signatures targeted by any survey are produced by the coupled effects of temperature, mineral alteration, porosity, permeability, saturation and fluid chemistry — so every measurement is sensitive to several of these at once. That is the root of non-uniqueness, and it plays out differently for each property.
Seismic velocity and attenuation are powerful but ambiguous. Increasing crack density, steam saturation and strong thermal gradients typically lower velocity and raise attenuation, while the Vp/Vs ratio can move either way depending on crack compliance, fluid phase and pore pressure. A low-velocity zone may mean elevated temperature, intense alteration, high porosity, high fracture density or partial melt. Its real value is dynamic: seismic data tells you where faults are active, where brittle failure concentrates, and how deformation and permeability evolve through time.
Electrical resistivity is often the single most diagnostic geothermal property because it responds strongly to alteration mineralogy and to the connectivity of saline fluids. The recurring pattern — a conductive smectite-rich clay cap over a more resistive productive interval, with deeper conductors tied to brines, partial melt or graphitic zones — is exactly why magnetotellurics (MT) has become a worldwide geothermal standard. But extremely low resistivity can equally mark saline brine or acid alteration, so resistivity alone is not fully diagnostic either.
GPR is frequently omitted from geothermal reviews because it sees only the shallowest metres. That omission is a mistake. Its strength is not reservoir-scale imaging but centimetre-to-decimetre resolution of the shallowest part of the system — where fractures daylight, alteration fronts migrate, shallow steam-heated zones develop and civil infrastructure meets unstable ground. In resistive volcanic lavas, welded tuffs and dry pyroclastics it delivers detail no deep method can match; in wet, clay-rich altered ground it can fail almost completely. Read correctly, that sensitivity to conductivity is information, not a defect.
The Review Approach
This is a narrative, problem-driven review rather than an attempt to catalogue every Mediterranean survey. Its aim is to make method complementarity explicit and auditable. Rather than assert that “integration is useful,” it converts the qualitative literature into a transparent comparison across five decision criteria that actually drive geothermal projects: mapping the clay cap and alteration, resolving active fault and fracture dynamics, characterising shallow flow and leakage, supporting 4D (time-lapse) monitoring, and delivering quantitative uncertainty.
Campi Flegrei was chosen as the anchor case because it combines geothermal interest, volcanic unrest, dense urban exposure and unusually rich multi-method observations — short-period MT imaging of the central caldera built from 47 independent soundings, 4D InSAR deformation modelling of the feeding system, and multidisciplinary shallow geothermal exploration in the eastern sector. That density allows a genuine test of what each method contributes, scored below.
The Campi Flegrei Synthesis
The value of Campi Flegrei is that it exposes both the strength and the non-uniqueness of every observable at once. A conductive anomaly may indicate smectite-rich alteration, saline fluids, acid-sulphate alteration or clay-rich sediment. A low-velocity zone may reflect hot fractured rock, pore-pressure effects or lithological change. Uplift may arise from pressure changes at very different depths and geometries. The key feature is not a single anomaly but the spatial coincidence and partial mismatch between deformation, seismicity, hydrothermal discharge and conductive structure.
Read together, the methods separate cleanly by role: MT gives the clearest image of the electrical architecture and the conductive, fluid-rich volumes; seismicity and attenuation track brittle failure, pressurisation and fracture activation; InSAR constrains the temporal deformation field but cannot by itself distinguish magmatic, hydrothermal or structural sources; and shallow geophysics (SP, TDEM, targeted GPR) governs the questions of shallow circulation, leakage, altered ground and site suitability. The table below is the synthesis in operational terms.
| Method | Main diagnostic contribution | Decision value | Main uncertainty / ambiguity |
|---|---|---|---|
| Short-period MT / MT | 3D resistivity architecture; conductive hydrothermal / alteration volumes; deeper roots of the system. | Best method for locating clay-cap / reservoir-scale electrical structure and prioritising zones for integrated interpretation. | Conductivity is non-unique: clays, brines, temperature, acid alteration and lithology can produce similar responses. Static shift, site spacing and regularisation affect geometry. |
| Passive seismic / tomography | Earthquake distribution, velocity / attenuation anomalies and brittle volume linked to pressurisation and active structures. | Best dynamic constraint on fault / fracture activation and unrest-related permeability change. | Requires adequate event / station geometry. Velocity anomalies are not uniquely geothermal and need petrophysical / geochemical calibration. |
| InSAR / geodesy | Time-dependent deformation and source migration in the caldera feeding system. | Essential for 4D monitoring and for separating persistent structure from transient pressure change. | Source models can be non-unique; deformation alone cannot distinguish magmatic, hydrothermal or structural mechanisms. |
| SP and TDEM | Shallow hydrothermal flow pathways, discharge zones and upper resistivity layering. | Bridge between surface observations and deeper MT / seismic interpretation; useful for environmental and shallow-resource decisions. | Sensitive to weather, electrode / reference choices, cultural noise and near-surface conductivity heterogeneity. |
| Targeted GPR | Very shallow fractures, altered ground, buried utilities and infrastructure-scale hazards in resistive sectors. | High-resolution tactical tool for shallow risk reduction and engineering planning. | Strong attenuation in wet clay-rich altered ground; limited direct reservoir value. |
Three conclusions follow, and they reshape how a Mediterranean programme should be designed. First, integration must be scale-specific: InSAR and seismicity for temporal unrest, MT / CSAMT for reservoir-scale electrical architecture, SP / TDEM / GPR for shallow hydrothermal and engineering questions. Second, the strongest interpretation is not the most detailed image but the one with the fewest unresolved alternatives after independent constraints are compared. Third, the same logic transfers as a template to other systems — weighting MT / CSAMT and SP for clay-cap and upflow mapping at Milos and Nisyros, 3D MT and seismic structure along active graben faults in western Anatolia, or seismic and deformation monitoring at Etna where fracture activation dominates the signal.
A Staged Monitoring & Exploration Workflow
These principles collapse into a practical, deliberately modular four-stage workflow. Not every field needs every method — but most failures in geothermal geophysics happen because surveys are chosen for convenience or habit rather than because they address the dominant uncertainty at the correct scale.
- Stage 1 — Regional reconnaissance. Broadband MT, structural geology, remote sensing and baseline seismicity identify overall architecture, main fault corridors and probable upflow zones. The goal is reduction of regional ambiguity, not detailed reservoir mapping.
- Stage 2 — Target refinement. CSAMT, focused seismic imaging, gravity or shallow EM sharpen the geometry of the upper reservoir, clay-cap boundaries and drilling targets. Even sparse borehole control becomes disproportionately valuable here, because it calibrates both the geophysics and the conceptual model.
- Stage 3 — Near-surface & infrastructure characterisation. GPR, SP, TDEM and engineering-scale surveys map shallow fractures, leakage pathways, thermally altered soils and buried utilities. In developed fields this stage is about risk management and environmental oversight, not just exploration.
- Stage 4 — Time-lapse monitoring. Repeat seismic, EM, SP and geodetic observations are designed around expected operational or natural change, with the emphasis on repeatability, uncertainty thresholds and integration with production or unrest data — not ad hoc re-surveys.
| Method family | Typical depth | Primary sensitivity | Main strengths | Main limitations |
|---|---|---|---|---|
| Local earthquake / passive seismic | 1–30 km | Vp, Vs, Vp/Vs, attenuation, event distribution | Tracks active structures, fluid pathways and time-dependent change | Needs good event / station geometry; non-unique petrophysical interpretation |
| Ambient-noise / nodal seismic | 0.5–10 km | Surface-wave velocity, temporal velocity change | Useful where natural seismicity is limited; repeatable for monitoring | Resolution depends on array geometry and noise field |
| Reflection / refraction seismic | 0.1–5 km | Acoustic impedance and interface geometry | High structural resolution where acquisition is feasible | Logistically demanding in rough volcanic terrain; strong scattering |
| MT | 0.1–10+ km | Bulk electrical resistivity | Best deep EM method; maps clay cap to deep roots | Diffuse resolution; static shift and inversion non-uniqueness |
| CSAMT | 20–1500 m | Shallow-to-intermediate resistivity | Higher resolution than MT in the target depth range | Near-field effects and cultural noise must be controlled |
| TDEM / TEM | 30–500 m | Transient resistivity response | Good bridge between shallow and deep EM | Depth penetration still limited in highly conductive ground |
| Self-potential (SP) | Surface, shallow flow | Electrokinetic and thermoelectric potentials | Directly sensitive to groundwater and hydrothermal flow | Sensitive to referencing, weather and conductivity heterogeneity |
| GPR | 0–30 m (more in resistive ground) | Dielectric contrasts, near-surface conductivity | Very high resolution; excellent for fractures and infrastructure | Strong attenuation in wet clay-rich altered ground |
Why This Matters Beyond Campi Flegrei
The Mediterranean’s real advantage is that it packs several contrasting geothermal system classes into a compact region — which makes it a comparative natural laboratory rather than just a collection of local case studies. Volcanic island systems such as Milos and Nisyros are ideal for testing the coupling between MT-defined conductive caps, shallow radar-visible structure and gas-rich discharge. Caldera systems like Campi Flegrei demand full integration of seismicity, EM imaging, hydrochemistry and geodesy, because the same signals matter for both development and hazard. Extensional continental systems in western Anatolia show how deep fault permeability and basin architecture control reservoirs without a large volcanic edifice. The monitoring recipe that works for one is transferable — but only if each method is reassigned to the dominant uncertainty of the new play type.
| Province / play type | Representative sites | Most diagnostic methods | Main monitoring priority |
|---|---|---|---|
| Active caldera systems | Campi Flegrei, Santorini | Passive seismic, MT, InSAR, SP | 4D unrest tracking; separating hydrothermal vs magmatic signals |
| Volcanic-island high-enthalpy | Milos, Nisyros, Vulcano, Ischia | MT / CSAMT, TDEM, SP, targeted GPR | Reservoir geometry, shallow leakage, structurally controlled permeability |
| Basaltic stratovolcano / flank | Etna | Seismic monitoring, InSAR, SP, local EM / GPR | Time-dependent fracture activation and hydrothermal circulation |
| Continental extensional | Gediz & Büyük Menderes grabens, W. Anatolia | 3D MT, seismic imaging, gravity, hydrochemistry | Target refinement along active fault corridors and basin margins |
| Low- / medium-enthalpy urban | Eastern Campi Flegrei, direct-use targets | Shallow seismic, gravity, TDEM, SP, selective GPR | Site screening, environmental control, infrastructure compatibility |
Where the field is heading
Four trends will shape the next decade: distributed fibre-optic sensing, which turns wells and existing fibre into dense acoustic, thermal and strain arrays; tighter coupling between subsurface geophysics and surface deformation, as 4D geodetic imaging links deformation time series directly to an evolving feeding system; lighter, more flexible shallow platforms — low-frequency large-depth radar, drone acquisition and rapid-deployment arrays — for steep or hazardous terrain; and machine learning, whose best near-term role is assisted interpretation and pattern extraction across multi-method time series, not black-box replacement of physics-based inversion.
The Review at a Glance
| Core thesis | No single method characterises a geothermal system; uncertainty-aware integration does. |
|---|---|
| Deep architecture | MT / CSAMT — clay caps, reservoirs, deep conductive roots. |
| Active structure & dynamics | Seismic methods — faults, fracture activation, time-dependent change. |
| Temporal deformation | InSAR / geodesy — 4D monitoring of the feeding system. |
| Shallow domain | SP, TDEM, GPR — leakage, altered ground, infrastructure, site suitability. |
| Anchor case study | Campi Flegrei caldera — 47 MT soundings, 4D InSAR, shallow multidisciplinary data. |
| Decision rule | Match the method to the dominant uncertainty; trust the interpretation that survives cross-method testing. |
| Recommended workflow | Reconnaissance → target refinement → near-surface / infrastructure → time-lapse monitoring. |
Frequently Asked Questions
Which geophysical method is best for geothermal exploration?
There is no single best method. Magnetotellurics (MT) is the cornerstone for deep exploration because electrical resistivity responds strongly to clay caps, saline fluids and hydrothermal alteration — but a conductive anomaly is non-unique, and MT is far less informative about active faulting or shallow leakage. Seismic methods resolve active structure and temporal change; SP, TDEM and GPR resolve the shallow domain. The right choice depends on which decision uncertainty dominates your project, which is exactly why the strongest programmes integrate several methods.
How deep can each method actually see?
As a rough guide: passive seismic tomography spans roughly 1–30 km; MT reaches 0.1 km to more than 10 km; CSAMT covers about 20–1500 m; TDEM about 30–500 m; and GPR typically images 0–30 m, occasionally more in resistive volcanic ground. SP and InSAR are surface methods sensitive to flow and deformation rather than a fixed depth. These ranges overlap only partly — which is the point of combining them.
Why does GPR appear in a geothermal review at all, given its shallow penetration?
Because the shallowest part of a geothermal system is where leakage, ground cracking, alteration fronts and infrastructure hazards first appear — and no deep method resolves that ground at decimetre scale. In resistive volcanic units, dry fractured lavas and pyroclastics, GPR is a high-resolution tactical tool for shallow fracture mapping, caprock discontinuities and pipe-corridor or subsidence work. Its main limit is strong attenuation in wet, clay-rich altered ground, so it should be deployed where the geology suits it.
Why focus specifically on the Mediterranean?
The Mediterranean geothermal belt packs several contrasting system classes — active calderas, volcanic islands, basaltic stratovolcanoes and continental extensional grabens — into a compact region. That makes it a comparative natural laboratory in which transferable monitoring strategies can be tested across different play types, rather than a single-analogue setting. Transferring exploration concepts from Icelandic, Japanese or Basin-and-Range analogues must always be done critically, with explicit attention to local geology.
What does a well-designed geothermal survey programme look like?
A practical, modular four-stage workflow: (1) regional reconnaissance with broadband MT, structural geology and baseline seismicity; (2) target refinement with CSAMT, focused seismic and shallow EM, calibrated against any available borehole control; (3) near-surface and infrastructure characterisation with GPR, SP and TDEM; and (4) time-lapse monitoring designed around expected operational or natural change. Not every field needs every stage — the aim is to address the dominant uncertainty at the correct scale.
Why does the review emphasise uncertainty so heavily?
Because most published geothermal interpretations present a preferred model without clearly separating what is directly constrained by data from what is inferred through conceptual reasoning. That is especially problematic for MT inversion in complex volcanic terrain and for seismic interpretation in heterogeneous near-surface settings. Reporting sensitivity, resolution and plausible alternatives explicitly is what moves geothermal geophysics from persuasive imaging to measurable risk reduction.
Planning geothermal exploration or monitoring?
Geomorph Imaging Solutions designs and interprets integrated geophysical programmes for geothermal, volcanic and infrastructure projects across Greece, Cyprus, Italy, western Anatolia and the wider Mediterranean. We are hardware-agnostic: the value sits in survey design, processing and uncertainty-aware interpretation — matching the method to the decision, not the logo to the antenna. Tell us about your prospect and we will tell you honestly what geophysics can, and cannot, resolve.
Discuss your project with us
Full publication: Arvanitis, M. (2026). “Integrated geophysical monitoring of active
geothermal systems in the Mediterranean: a review of seismic, magnetotelluric, CSAMT and near-surface EM/GPR
methods.” GeoStudies. Open access under CC BY 4.0.
Figures 1 and 2 are review-level syntheses reproduced from the above.
ORCID: 0009-0003-6085-7276
Related:
Finding the Basilica of Ancient Messene — GPR before excavation ·
Our geophysical survey services