GPR for underground void locating

[topgpr]

Underground void locating with GPR is not always easy. Sometimes it's just hard to tell whether you are looking at a void or not in the GPR profile. In the following example,between 5' to 9' marks, there is apparently a void since it's so close to the ground surface. But it's harder to determine whether the anomaly around 13' to 14' marks at depth about 2' is a void not. Don't know whether there are some data processing tricks to help in this kind of situation.

[geophix]

There would be some techniques for enhancing the visual effects of GPR profiles such as filtering, phase analysis or even migration. But ultimately it depends on the human to make the final interpretation. Maybe artificial intelligence would be the way to go for the best help in the future.

[Tyler]

GPR data simulation could be helpful, too.

[geophix]

I actually did a GPR data simulation for underground void locating long time ago. As shown below, the ringing effect can be seen when the space between top and bottom of the void is small, I guess it's due to the bouncing waves between the two interfaces. When the space is large, you won't see the ringing effect. But no matter whether the space is small or large, the reflections on top of voids are usually strong.

[topgpr]

Thanks, guys. Wonderful data example, geophix!

[geophix]

Here is a GPR data example of underground voids. It looks like the void to the right shows multiples within the void, so the void might not be that big depth wise.