Time-depth conversion issue

[Kurt Soncco]

Hello everyone, 

First, thanks to all the support about this amazing package for GPR modelling. 

I'm having issues with the time-depth conversion. I'm performing A-scans of different frequencies of the same model: A wood member of 0.50 m of depth and a steel plate at the bottom. Check geometry in the attached file, as well as the input file I used (I only change the frequency of the wave).

The problem is that, calculating the velocity of the wave in the medium, I figured that the velocity does not change as the frequency increases. So, my question and problem are: why A scans changes as the frequency increases? And why after I calculate the depth using this formula the peaks of the wave do not appear to be in the position of the plate?

v=\frac{c}{\sqrt{\varepsilon}}; z=\frac{v\times t}{2}

In my research group, we expect that the peaks will appear in the same depth positions, as the geometry and velocity does not change. 

I'm attaching some plots I have for several frequencies and how my time-depth conversion looks like. 

Thank you for the time of reading the questions,
Kurt Soncco
Research Assistant IDEA LAB
University of Illinois at Urbana Champaign

[Craig Warren]

Dear Kurt,

Changing the centre frequency of the waveform will change the frequency content, but importantly will also effect the onset of the waveform - see, for example, the definition at http://docs.gprmax.com/en/latest/plotting.html#gaussiandot

If you perform time-zero corrections on your A-scan you should find that the reflections 'line up'.

Kind regards,

Craig