Time Domain/Transient Electro-magnetic (TDEM)

[pepperj]

Reading this would be above my pay grade.

Electromagnetic Methods in Applied Geophysics

Volume 2 covers, in depth, the physical basis of EM methods of exploration magnetometric resistivity method, profiling methods using small sources, large-layout harmonic field systems, EM soundings, time-domain EM prospecting methods, VLF, MT, CSAMT, airborne EM methods, borehole EM techniques...

books.google.ca

 

[geowizard]

What you are showing here appears to be the raw TDEM decay data, not the inverted data which is what geophysicists would provide as a deliverable. We would apply a series of inverse models such as laterally constrained inversion (LCI) and then spatially constrained inversion (SCI) to the raw decay data to generate 2D profiles of electrical conductivity versus depth. The SCI data can then be interpolated into 3D conductivity volumes.

Understood.
The question is, "Where does science become art?" I can share examples where SkyTEM did inversions that took two weeks and crashed. Inversions that showed excellent conductors that had no correlation with physical drill holes. The drill holes were consistently "no shows". To be fair, there is the possibility that the holes missed their intended targets. The client company walked away.

The challenge for the analyst is that TEM anomalies don't provide qualitative information in terms of I-Q and phase. Phase has been shown by example in FDEM to provide discrimination - adding a qualitative dimension to targets. This dimension reduces risk in moving forward in exploration of precious metals.

- Geowizard

 

[geowizard]

I don't think you understand that electromagnetic induction is a diffusion equation governed system. At frequencies up to low or mid kHz there is not a time/speed/distance calculation. Only at higher frequencies when the electromagnetic systems transitions to wave equations do time/speed/distance matter. Ground penetrating radar (GPR) operates from mid kHz range to low GHz frequency ranges. At lower frequencies where metal detectors or TDEM systems operate it is a diffusion equation.

I understand.

Fundamentally, TDEM is a function of dv/dt and dB/dt which by definition include time. We know the functions dB/dt and dv/dt (decay time) are the response relative to T0. The electromagnetic "response" in the medium is in the Time domain. You are correct.

Travel Time for penetration through the medium is not instantaneous but trivial.

ed: added "trivial"

- Geowizard

 

[geowizard]

What you are showing here appears to be the raw TDEM decay data, not the inverted data which is what geophysicists would provide as a deliverable. We would apply a series of inverse models such as laterally constrained inversion (LCI) and then spatially constrained inversion (SCI) to the raw decay data to generate 2D profiles of electrical conductivity versus depth. The SCI data can then be interpolated into 3D conductivity volumes.

My comment on Inversion;

There is a point where science becomes art. Raw data in terms of a real TDEM response is as valid as the data can get. Filtering, smoothing and inversions and doing a best fit perturb the data into art - not science.

The UBC.edu document summarizes six methods of obtaining tau.

SkyTEM openly admits inversions that require over 100 hours of number crunching and then crash.

https://www.tau-geophysics.ca/resources/Geosyntesis%202011_Decay%20Constants%20in%20AEM.pdf

Show me a response. I will do the rest!

- Geowizard

 

[k0bgalan]

Lots of instruments use time domain measurements. The one I'm most knowledgeable about is the TDR—Time-Domain Reflectometer. The one I have is used to detect anomalies (breaks, opens, shorts, etc.) in telephone cabling. It can even tell you where, if any, the loads are. Not cheap, but a lot cheaper than digging a huge trench!

And... All of this makes me wonder about Garrett's VX (Vortex) series. It is, according to Garrett, the only fully-digital hobby metal detector they make. I've also been told, the only time the depth measurement is accurate (no + or - percentage given) is when the built-in pin pointer is used—a fact I can attest to. I would tend to believe it uses a form of time domain measuring for this function?