Transient electromagnetic surveying, usually abbreviated TEM or TDEM, investigates underground electrical conductivity by measuring a magnetic response after a controlled transmitter current changes or switches off. If you are reading a time-domain electromagnetic decay curve, late-time voltage gate, resistivity sounding or conductive-layer inversion, start by identifying the loop geometry, waveform and actual measured channel. A decaying signal is not a direct image of a rock layer, and a late time gate does not correspond to one exact depth.
A transmitter loop produces a magnetic field whose change induces electrical currents in the ground. After the transmitter is turned off, those currents decay and redistribute, while the receiver detects a time-dependent secondary response. The shape of that response contains information about electrical properties beneath and around the loop. A model is required to infer a conductivity or resistivity profile, and multiple models can sometimes fit the same noisy decay curve.
This guide explains TEM data from current waveform and receiver geometry through switch-off transients, early and late time gates, noise floors, apparent resistivity, one-dimensional inversion, resolution and depth of investigation. It includes independently checkable illustrative calculations with units and assumptions. Its purpose is to help readers distinguish an observed voltage decay from a modelled underground layer, while recognising the method’s value for hydrogeology, engineering and environmental exploration.
Quick answer: what is a TEM sounding?
A TEM sounding is a set of time-dependent electromagnetic responses recorded at a survey location using a stated transmitter–receiver arrangement. An instrument may report induced receiver voltage, a normalised field response or another processed signal across discrete time windows after the transmitter-current transition. Analysts compare those measurements with predictions from electrical-earth models.
In many applications the result is an inferred vertical sequence of resistive and conductive intervals. Those layers are interpretations derived from the decay curve and modelling assumptions, not direct material samples. A reliable report describes both the observed time-domain response and the resistivity model, including where the data no longer constrain depth reliably.
Time-domain is not a calendar measurement
The word time in TEM refers to milliseconds or other intervals after a controlled transmitter waveform transition, often a current shut-off. It does not ordinarily mean a geological history through thousands of years, or a comparison of the ground at two seasons. A time gate samples the instrument’s transient electromagnetic response over a defined interval.
Misreading the term can distort the entire interpretation. Time-domain measurements can be repeated on different dates, but that is a separate time-lapse study. First identify the transient time axis and its units before deciding whether a graph describes spatial variation, electrical response over microseconds or actual environmental change across months.
How a transmitter loop creates a changing field
A conductor arranged in a loop carries a controlled electrical current during a TEM measurement cycle. The current generates a magnetic field. When the current changes, electromagnetic induction produces electric fields and currents in nearby conductive material. The earth’s induced currents generate a secondary magnetic field that evolves after the transmitter field is changed.
The US EPA overview of transient electromagnetic methods describes this basic mechanism. It does not suggest that the instrument sends a narrow light-like ray to a selected depth. TEM sensitivity is distributed through a three-dimensional volume and depends on loop size, electrical structure and time since the current transition.
Why currents are called eddy currents
An induced electric field can make electrical current circulate through conductive material. Such currents are often described as eddy currents because they form loops within the conducting medium. Their distribution evolves as the electromagnetic field diffuses through the ground and energy dissipates. This is not fluid water moving in circular eddies, although the same word is used.
A source-to-receiver diagram may show idealised rings expanding or moving deeper with time. That is a conceptual drawing of electromagnetic diffusion under assumptions, not a camera image of discrete current circles. Real electrical structures can steer or distort current paths. A labelled ring in a textbook should not be interpreted as a precise underground boundary detected by the instrument.
Why the receiver measures after transmitter shut-off
The transmitted primary field can be much stronger than the subtle secondary response of interest. Time-domain measurements are often collected after the transmitter’s current has been switched off, within carefully controlled intervals. This helps isolate the decaying response associated with induced currents in the ground.
However, current does not necessarily stop instantaneously. The finite transmitter turn-off waveform, circuit characteristics and receiver recovery affect the earliest useful measurements. A method that assumes perfect instantaneous switch-off when it was not achieved can distort early-time inversion. The actual waveform is part of the measurement, not an irrelevant equipment detail.
What is a time gate?
A time gate is a defined interval after the relevant transmitter transition over which a receiver signal is sampled, integrated or averaged. Instead of recording one infinitely precise instant, instruments often use windows designed to balance temporal resolution and signal-to-noise performance. Gates may widen at later times as the signal becomes weaker.
A gate labelled 0.5 milliseconds might represent the centre of a window rather than its beginning or end. The method should define the gate timing convention. Comparing two surveys whose nominal gate labels differ slightly requires care, especially where early-time response changes rapidly. The time stamp alone does not tell the whole sampling procedure.
Early time and late time are relative descriptions
In a common TEM sounding, early-time measurements are often more influenced by shallower or faster-changing responses, while later times can contain information from deeper or slower-decaying conductive structures. This relation is useful but broad. The actual sensitivity depends on conductivity distribution, source size, geometry, waveform and noise.
It would be incorrect to call a 1-millisecond gate the ‘one-metre depth channel’. The instrument does not have a universal time-to-depth ruler. A late-time response can also be dominated by a shallow, very conductive layer or by noise. Depth of investigation must be estimated through the survey model and its sensitivity rather than through time labels alone.
Why a conductive medium may decay differently
Conductive subsurface materials can sustain induced currents differently from more resistive materials. In a simplified comparison, conductive structures often allow a stronger or more slowly decaying secondary response, while highly resistive media can yield rapid attenuation. But the measured transient is the combined effect of conductive distribution and source geometry.
One cannot assign a particular soil type from decay slope alone. Clay minerals, pore-fluid salinity, saturation and conductive mineral phases can all affect bulk electrical properties. The same broad decay shape can sometimes be explained by alternative layer thicknesses and conductivity contrasts. A physical interpretation needs complementary geological evidence.
A decay plot’s axes must be read first
The horizontal axis may show time after transmitter turn-off in microseconds, milliseconds or seconds, sometimes on a logarithmic scale. The vertical axis can show induced voltage, magnetic-field derivative, normalised response or a transformed apparent resistivity. These are different physical quantities. A curve rising on one transformed axis may correspond to a falling raw voltage magnitude.
Identify sign conventions and whether absolute values, logarithms or filtering were applied. A negative voltage is not automatically proof of a negative conductivity material. It may reflect loop orientation, receiver polarity, waveform convention or noise. Ask what the instrument actually recorded before offering a story about the subsurface.
Logarithmic time axes can mislead intuition
A logarithmic axis gives equal visual spacing to multiplicative time changes. The distance between 0.1 and 1 millisecond may look similar to the distance between 1 and 10 milliseconds, even though the second interval is nine milliseconds long and the first is only 0.9 millisecond. A steep-looking curve is not necessarily changing linearly per second.
When comparing gates, convert their labels to consistent units and inspect the actual time differences. Do not infer a rate from screen distance unless the scale is understood. This is particularly important in TEM, where many gates span orders of magnitude and late values can be close to measurement noise.
What does the receiver coil observe?
The voltage induced in a receiver coil relates to the time variation of magnetic flux through its area, under Faraday’s law. The recorded amplitude can depend on coil turns, effective area, orientation and instrument gain. Some equipment displays calibrated magnetic-field derivatives instead of raw voltage.
A plot labelled voltage in nanovolts or microvolts cannot be converted to a material conductivity simply by deleting its unit. The transformation requires instrument calibration and a physical response model. A report should identify the receiver’s measured channel and normalisation before comparing amplitudes from different equipment.
Loop area affects the response
For a simple square transmitter loop with side length 20 metres, the enclosed area is 400 square metres. A square with side length 40 metres encloses 1,600 square metres, four times as much area. The geometry can influence the transmitter moment and the sensitivity of the measurement, subject to current, turns and actual field configuration.
It would be wrong to conclude automatically that the 40-metre loop measures four times deeper or produces exactly four times every recorded voltage. Loop area affects one aspect of source geometry; depth response also depends on subsurface conductivity, waveform, receiver placement and noise. The numerical area calculation is useful only when accompanied by those distinctions.
Central-loop and offset-loop configurations
TEM surveys may use receiver loops located centrally within the transmitter, at different offsets or in other arrangements designed for particular targets. Changing the receiver position changes how the secondary electromagnetic field is sampled. Two surveys can therefore respond differently to the same underground structure even when their time gates are identical.
A cross-survey comparison should record loop dimensions, turns, current, receiver position and orientation. A published model from one configuration may not be directly transferable to another. The survey geometry is part of the input to forward modelling and inversion; it should not be reduced to a line reading ‘TEM acquired at the site’.
Current waveform and shut-off time
The transmitter current may be switched in a periodic waveform rather than supplied as a single isolated event. The timing, ramp duration, repetition rate and cycle polarity can affect the recorded signal and opportunities for averaging. Different TEM systems may process positive and negative half-cycles to reduce some instrumental or environmental noise.
Read the instrument specification and actual acquisition settings. If the transmitter turn-off is slow relative to the earliest intended gates, a simplistic instantaneous model may be inappropriate. A credible paper explains how waveform effects were included in processing or which early gates were excluded. A high-quality-looking decay curve is not proof the assumed waveform was accurate.
Why stacking repeated transients helps
A field system can repeat the transmitter cycle many times and combine receiver measurements to reduce some uncorrelated noise. Under idealised independent-noise conditions, averaging N comparable measurements can improve signal-to-noise ratio roughly with the square root of N. Persistent interference, systematic drift and coherent sources do not vanish by the same rule.
For example, sixteen independent repeats would yield an idealised factor four in signal-to-noise improvement, not sixteen. Real field gains may be smaller. A report should say how many stacks or averages were used and whether later data passed noise or coherence checks. The number of repeated pulses is not the number of independent geological samples.
Very early gates can be instrument-limited
The earliest time gates may contain significant contributions from transmitter turn-off, receiver recovery or electronic transients. Some systems have an effective dead time during which reliable secondary-field data cannot be acquired. Attempting to interpret those samples as shallow geology can inject false structure into an inversion.
A practical report marks the valid measurement window. If an algorithm fits implausibly high or low resistivity near the surface while early gates are noisy, inspect whether the waveform and recovery were handled correctly. Excluding unreliable gates may improve physical interpretation, but the exclusion rule should be documented rather than chosen solely to make a model look smooth.
Later gates are often limited by noise
At late times the useful secondary response can become weak, especially where the electrical structure is relatively resistive. External electromagnetic noise, motion, small instrument offsets and numerical precision can then dominate the apparent signal. A noisy late tail is not evidence for an infinitely deep or extremely conductive layer.
The analyst should identify gates above the effective noise floor and distinguish them from those retained only as weak constraints. If an inversion forces a smooth fit through random late-time fluctuations, the resulting deep structure may be an artefact. Depth estimates need to account for signal-to-noise and sensitivity, not just the longest nominal time recorded.
Noise from power lines and other infrastructure
Alternating-current electrical infrastructure can generate electromagnetic interference. Nearby fences, large metal structures, motors and communication systems may also influence recorded transients. In urban environments, some data windows can be more affected than others depending on interference frequency and acquisition timing.
A responsible field report notes known sources and quality-control procedures. Repeating a sounding under changed conditions can help test whether a suspicious late-time pattern is reproducible. However, a repeated interference source can also be coherent. Agreement between repeated traces is valuable but does not automatically prove the signal originates in underground geology.
Why negative or oscillating gates need context
Some TEM datasets contain negative-valued measurements because of polarity conventions, instrument response, sign changes in secondary fields or noise. A sign reversal might be physically meaningful in particular complex models or configurations, but it should not be declared a negative conductivity. Conductivity and resistivity in ordinary passive earth materials are not inferred as negative simply because a voltage sample has a minus sign.
Check whether the plotted value is signed, absolute, transformed or noise-corrected. On logarithmic plots, negative measurements may be omitted or treated specially. A missing plotted point is not necessarily a missing acquisition sample; it may have been excluded by display rules. Interpret the actual data file and quality flags, not just the appearance of the curve.
Normalisation affects cross-instrument comparisons
One system may display receiver voltage, another voltage divided by transmitter current or effective receiver area, and another an apparent-resistivity transformation. Different normalisation choices can make identical physical responses appear numerically different. Comparing raw values without their units and configuration can produce erroneous field conclusions.
A good report states what the ordinate means, which calibration was applied and how instrument settings were handled. A plotted curve with arbitrary units can still support qualitative within-survey comparisons under stable conditions, but it should not be used to claim exact conductivity without a documented forward model.
Apparent resistivity is a processed representation
Some TEM software transforms the measured decay response into an apparent resistivity under a reference model and defined formula. This creates a convenient display for comparing changes through time, but the result is not the intrinsic resistivity at one precise depth. A plotted apparent-resistivity curve may vary across gates because the electromagnetic sensitivity shifts, not because the instrument directly sampled a succession of thin underground layers.
The transformation also depends on geometry and signal convention. A reader should not compare an apparent-resistivity curve with raw receiver voltage values as if they were identical measurement scales. First identify whether the paper displays the primary observation, a transformed response or the output of an inversion. Each stage supports a different kind of inference.
Resistivity and conductivity are reciprocals in the simple model
For a uniform isotropic material, resistivity in ohm-metres equals one divided by conductivity in siemens per metre. A modelled layer of 10 ohm-metres therefore corresponds to conductivity 0.1 S/m. A modelled layer of 100 ohm-metres corresponds to 0.01 S/m. This arithmetic is straightforward; the challenge is determining whether the modelled layers accurately represent the heterogeneous earth.
A report that changes from resistivity to conductivity must carry the correct units and clarify whether values are apparent or intrinsic estimates. Comparing raw voltage decay with conductivity is not a mere unit conversion. A forward electromagnetic model connects the material properties to the measured transient response.
Why an inversion is needed
An inversion is a procedure for finding one or more earth models whose predicted electromagnetic responses resemble the observed data. The forward model calculates what a hypothetical conductivity distribution would produce for the stated loop geometry and waveform. The inversion adjusts model parameters in search of acceptable agreement.
This is not the same as directly observing each subsurface layer. Different conductivity structures may fit similar transient data, particularly when noise, limited time windows or strong conductive barriers reduce sensitivity. A responsible interpretation shows the observed decay, modelled response and uncertainty before claiming an exact layer thickness or salinity distribution.
A one-dimensional layered-earth sounding
A common TEM interpretation assumes that conductivity changes mainly with depth beneath a sounding location. The resulting one-dimensional model consists of layers with estimated resistivities and thicknesses. Multiple such models may be assembled into a cross-sectional display along a survey line.
If the subsurface varies strongly sideways, a 1D sounding may absorb lateral effects into misleading vertical layers. Putting several 1D models beside each other does not magically create a fully three-dimensional inversion. A coloured transect built from soundings remains an interpretation whose assumptions should be described. The distinction matters around buried channels, vertical conductors and complex infrastructure.
Sharp and smooth inversions answer different questions
A sharp-layer model may represent the ground as a small number of interfaces separating relatively constant resistivities. A smooth inversion allows conductivity to vary gradually with depth using many model cells. Both can fit observations under different regularisation choices and constraints.
The visually sharper display is not automatically closer to reality. Smooth models may smear a genuine boundary, while blocky models may invent a discrete interface to approximate a gradual transition. Compare the model-data fit and the geological setting. Reporting only the preferred coloured cross-section hides the range of models supported by the observations.
Regularisation keeps an inverse problem stable
Because electromagnetic inversions can be nonunique and noisy, algorithms commonly impose constraints that favour smoothness, limited changes or plausible parameter ranges. These are regularisation choices, not new measurements of geology. They prevent unstable fits but also influence the resulting image.
A reader should ask which assumptions were used and whether a different reasonable setting changes the key geological claim. If an apparent conductive band appears only under one heavily smoothed inversion, confidence should be lower. A feature supported across several plausible settings and independent measurements is more persuasive, though still model-dependent.
Fit residuals are evidence about the model
After inversion, predicted responses can be compared with measured time-gate values. The difference is a residual. A good fit indicates that the chosen model reproduces the observations within a stated error model. It does not establish that this model is unique or that every inferred layer has been observed directly.
Inspect residuals across early and late gates. If discrepancies cluster in early times, the transmitter waveform or shallow structure may be misrepresented. If late-time residuals dominate, noise or deeper model limitations may matter. A single overall fit score can conceal systematic mismatches that change the geological interpretation.
Uncertainty weighting matters when signal becomes small
A decay curve may span orders of magnitude in response amplitude. If all gates are weighted equally in raw voltage units, large early-time signals can dominate the fit. If the inversion assigns unrealistic relative errors to weak late gates, it may overfit noise and produce unstable deep resistivity estimates.
Quality control should define a sensible noise model, gate exclusions and any floors applied to uncertainty. Readers need not solve the entire inverse problem to evaluate a report; they can ask whether the residuals were compared with measurement uncertainty and whether weak gates were treated cautiously.
Time does not convert directly to layer depth
A later gate often carries sensitivity to different and potentially deeper current distributions, but the relationship depends on conductivity, loop size and geometry. One cannot multiply the gate time by a universal wave speed and divide by two. TEM is an electromagnetic diffusion measurement, not the same travel-time problem as seismic reflection or ground-penetrating radar.
A simple example makes this distinction clear. Gates at 0.2 and 2 milliseconds differ by a factor ten in time. They do not necessarily correspond to target depths differing by a factor ten. The diffusion response must be modelled using actual electrical properties. A time label is an observation coordinate, not a depth ruler.
What depth of investigation really means
Depth of investigation describes the region where the available measurements meaningfully constrain the earth model under stated conditions. It depends on source geometry, conductivity structure, usable late-time data, noise and inversion assumptions. A nominal system specification such as ‘can sound to hundreds of metres’ is not a guarantee for every survey.
A shallow conductive layer can screen sensitivity to deeper resistive units. Conversely a favourable resistive setting and strong source moment may permit deeper information, subject to noise. A well-made report estimates where model values stop being data-supported. Colours displayed below that boundary should not be discussed as confirmed stratigraphy.
Conductive cover can hide deeper structures
Imagine a shallow conductive layer above more resistive rock. Induced currents may be strongly affected by the upper layer, limiting the response’s sensitivity to a deeper target. The inversion may fit many possible deeper resistivities with little change to the measured decay. In that case, the inability to resolve depth is a genuine property of the experiment.
It is misleading to present a crisp lower boundary solely because the software needs to draw a finite model. A colour band below the constrained region might be a starting assumption or regularised continuation. Ask whether data sensitivity, alternative models and independent logs support it before attaching geological names.
A resistive target is not always easy to detect
TEM often responds effectively to conductive bodies within relatively resistive surroundings, but a resistive body embedded in conductive ground can be difficult to isolate. The electromagnetic signal may be dominated by currents in the conductive host. Thus the technique is not an all-purpose void detector or universal direct map of every buried structure.
The EPA technical description of TEM describes this contrast dependence. Detection also changes with target dimensions, depth and loop geometry. A non-detection cannot establish that an unwanted underground object is absent, particularly when the object has a weak electrical contrast.
Conductive anomalies do not identify contaminants uniquely
A lower resistivity model zone could arise from saline pore water, clay minerals, saturation, conductive mineralisation or other subsurface changes. A contaminant plume is one possible interpretation in a relevant environmental context but not a chemical result of a TEM instrument.
To attribute a specific contaminant, collect fluid or soil samples and compare with laboratory chemistry, borehole logs and site history. When such evidence is unavailable, describe a conductive zone with possible explanations. The measured transient can identify where investigation is useful without performing chemical speciation.
TEM and FDEM use different parts of electromagnetic behaviour
Frequency-domain electromagnetic surveying uses an oscillating transmitter field and analyses phase-sensitive responses such as quadrature and in-phase components. TEM uses a transient field and examines signal decay after the transmitter transition. Both sense electrical properties through induction, but their acquisition geometry and inverse modelling differ.
The companion FDEM apparent-conductivity guide explains how phase and coil configuration influence a field map. Its low-induction-number assumptions should not be inserted unchanged into a TEM inversion. A current-switching response and a sinusoidal quadrature response are different observables.
TEM and ERT may complement each other
Electrical resistivity tomography injects current through ground-contact electrodes and measures voltage differences, whereas TEM uses changing magnetic fields to induce currents without the same electrode contact. Both can help characterise conductivity contrasts, but their depth sensitivity, spatial coverage and logistical constraints differ.
The EPA electrical resistivity guidance explains electrode arrays and inversion nonuniqueness. Agreement between ERT and TEM can increase confidence in a conductive-zone hypothesis when the methods are independently controlled. Yet each method may sample different volumes, so numerical model agreement should be examined rather than assumed.
A contrasting seismic measurement
Seismic refraction and reflection use elastic wave propagation, not electromagnetic induction. Their travel times depend on seismic velocities and wave geometry; TEM signal decay depends on electrical properties and diffusion. Thus a seismic layer boundary and a TEM conductive boundary need not coincide.
The seismic reflection CMP guide and surface refraction guide illustrate different axes, units and inverse assumptions. A multi-method investigation is most useful when these physical differences are preserved rather than treating all colourful underground sections as equivalent pictures.
A time-gate example for quality control
Suppose an illustrative receiver records positive decay magnitudes of 100, 20 and 4 arbitrary calibrated units at three increasing gates under one setup. Those numbers indicate decreasing recorded response magnitude through the selected times, not a direct sequence of soil conductivity values. Without the actual geometry, waveform and instrument calibration, the curve cannot be inverted quantitatively.
If the late measurement has uncertainty plus or minus 5 of the same units, the value 4 is not a reliable positive signal separated from noise. Giving it equal geological importance as the earlier values would be questionable. The diagnostic is to compare each gate to its uncertainty before asking the inversion to explain the tail.
A unit conversion exercise for a modelled conductor
An inversion proposes a layer resistivity of 25 ohm-metres. Under the homogeneous reciprocal relation, conductivity is 1 divided by 25, or 0.040 S/m, equivalent to 40 mS/m. A neighbouring modelled layer at 100 ohm-metres corresponds to 10 mS/m. The first is four times as conductive in this idealised comparison.
Does that prove the first layer contains four times as much salt or water? No. The resistivity estimates depend on the inversion, and bulk conductivity responds to multiple material properties. A modelled contrast gives a reason to test a soil or groundwater hypothesis, not a direct laboratory measurement of a constituent.
Repeat soundings and model stability
If a stationary TEM survey is repeated with the same loop configuration and processing settings, comparable time-gate responses provide some evidence of repeatability. But common systematic errors can repeat too, including environmental interference or a misrecorded loop dimension. A repeated curve does not prove the chosen inverted geological model is unique.
When possible, compare raw responses, noise estimates and model fits, not merely final colour profiles. Test whether plausible alternative starting models change the deep structure. A field survey becomes more robust when it documents those checks instead of selecting the inversion that looks most geologically attractive.
What to do with a suspicious late-time tail
Imagine a curve that behaves smoothly at early times but oscillates around zero at the last few gates. Rather than interpreting every crossing as alternating conductive and resistive layers, inspect the noise floor, interference, receiver stability and gate quality flags. A useful model may exclude late-time points not reliably distinguished from noise.
The exclusion should be documented and justified. A report can show accepted gate ranges and compare data with forward predictions. This makes the final inferred depth more honest: the survey may be informative to some depth without resolving every feature suggested by software outside that constrained interval.
Interpreting a profile built from separate soundings
Field teams may measure TEM at stations along a line and invert each sounding separately. When those one-dimensional models are placed side by side, the resulting display resembles a vertical cross-section. It can reveal spatial trends but may also show artificial lateral boundaries caused by station spacing or independent regularisation choices.
If a sharp wall appears between two widely spaced sounding positions, question whether the data truly resolve its position. A two-dimensional or three-dimensional model may better represent lateral effects, where appropriate. The safest description distinguishes between measured station locations and interpolated or inferred boundaries.
Independent transfer: what changed between two surveys?
Suppose a study reports stronger late-time response after a field treatment. Can we conclude that more groundwater appeared at a fixed depth? Not immediately. Changes in calibration, loop geometry, interference and ground conductivity can alter the response. The same signal difference may correspond to several electrical structures.
The correct procedure is to compare geometries and valid gate ranges, then test geological or hydrological models. Independent monitoring wells or water chemistry may strengthen one explanation. A good conclusion states which electromagnetic observation changed and how it is consistent with a particular environmental process under a declared interpretation.
Common interpretation mistakes
The first mistake is treating each gate as one exact depth; repair it by using sensitivity and inversion analysis. The second is treating receiver voltage as apparent resistivity without a model. The third is confusing a transient response with a propagating seismic wave. Their remedies begin with correct axes, physical principles and units.
Other mistakes include interpreting late-time noise as deep geological structure, assuming all conductive zones are contaminant plumes and treating software regularisation as direct observation. Repair these by inspecting error models, alternative conductivity structures, independent boreholes and the documented measurement chain. Strong conclusions arise from combined evidence, not from model colours alone.
Questions readers frequently ask
What is a TEM time gate? A defined sampling interval of the transient response. Does a later gate always represent one deeper layer? No. Does TEM require electrodes inserted into soil? Usually no, because induction provides coupling. Is the decay curve a direct conductivity profile? No. Can a transmitter turn-off affect early data? Yes.
Can interference affect late times? Yes. Can several earth models fit the same curve? Yes. Are conductivity and resistivity related reciprocally? Under a compatible homogeneous model, yes. Does a conductive anomaly prove water contamination? No. Is FDEM identical to TEM? No. Can TEM help guide borehole placement? Yes, while retaining model and depth uncertainty.
Authoritative method sources and further reading
The US EPA technical guide to time-domain electromagnetic methods explains current switching, eddy currents and decay-based inversion. Its companion TEM survey guidance describes configurations, applications and depth limits. The EPA overview of electromagnetic methods helps distinguish time-domain induction from FDEM and radar.
To keep the terminology clear, compare the FDEM phase and apparent-conductivity guide, seismic reflection CMP guide and cross-linguistic transfer guide. Those pages address different sources of apparent subsurface information and different risks of mistaking technical words for their everyday meanings.
Final lesson: the decay is observed, the layer is inferred
TEM measurements are valuable because the time-dependent electromagnetic response can constrain electrical-property variations beneath a survey location. But the modelled resistivity profile is an inference involving transmitter geometry, receiver calibration, usable time gates, assumptions about the earth and numerical inversion.
An accurate explanation keeps the chain in view: current waveform, induced earth currents, measured voltage decay, noise treatment, forward modelling, model fit and geological interpretation. Once those stages are separated, a reader can appreciate the method’s strength without mistaking a smooth inverted layer for something directly sampled by the instrument.