Skip to main content

Lyceum 2021 | Together Towards Tomorrow

This presentation demonstrates how UAV magnetic data is processed using Oasis montaj workflows

to locate and map buried targets such as service utilities. We also apply Euler deconvolution to estimate the target depths from a test site.

Overview

Speakers

Victor Mapuranga
Customer Solutions Specialist, Seequent

Duration

18 min

See more Lyceum content

Lyceum 2021

Video transcript

[00:00:00.217]
(ethereal music)

[00:00:10.600]
<v Victor>Thank you for dialing in today.</v>

[00:00:13.830]
I hope you have enjoyed Lyceum thus far.

[00:00:16.990]
My name is Victor,

[00:00:17.850]
and I’m a customer solution specialist at Seequent.

[00:00:21.480]
The title of my presentation is Euler

[00:00:24.177]
:a classic beat that still to the depths of geophysics.

[00:00:28.810]
I would like to acknowledge as pH engineering

[00:00:31.070]
we have shared their magnetic data acquired over a test site

[00:00:34.877]
at Bellow in Latvia,

[00:00:37.020]
using the synthesized MagDrone R3.

[00:00:42.647]
The test site consists of a number of buried ferrous

[00:00:45.380]
and non-ferrous targets.

[00:00:47.550]
Examples of which you can see in front of your screen

[00:00:51.360]
such as pipes and barrels.

[00:00:54.760]
The purpose of this presentation

[00:00:57.050]
is to show some tools that are available

[00:00:58.960]
in Oasis montaj that can be used

[00:01:01.240]
to estimate the depth of these objects.

[00:01:04.420]
Today we are going to focus on Euler deconvolution.

[00:01:09.780]
This is applicable in mapping objects such as utilities.

[00:01:14.080]
You can also use this method to estimate

[00:01:16.210]
the depth of unexploded ordinance.

[00:01:19.600]
We’ll be going through some workflows

[00:01:21.320]
for the estimation of depth for these types of objects.

[00:01:28.260]
Perhaps as a form of introduction

[00:01:31.030]
let us go over the Euler homogeneity equation

[00:01:35.460]
x, y as well as z is a location

[00:01:41.440]
or measurement point so to say.

[00:01:46.770]
X naught, y naught are know from our pre-target locations

[00:01:52.210]
and we are solving for z naught

[00:01:54.710]
which is the depth of the object.

[00:01:58.850]
The derivatives delta t over delta x,

[00:02:03.030]
delta t over delta y, and delta t over delta z

[00:02:07.660]
represent the gradients.

[00:02:11.090]
If you are using a total field magnetometer

[00:02:14.320]
you can derive these gradients.

[00:02:17.870]
If you are using a flux gate magnetometer

[00:02:20.610]
or multiple total field magnetometers in an array

[00:02:25.770]
one or all of these gradients can be measured directly.

[00:02:32.280]
And then t represents the measured field.

[00:02:37.610]
And b represents our background field

[00:02:41.630]
we call N our structural index

[00:02:44.370]
and it accounts for the rate of decrease

[00:02:46.380]
in the amplitude of the response

[00:02:48.050]
with distance from the source.

[00:02:50.830]
It ranges from 0 to three for potential field data.

[00:02:56.050]
All right, let’s go over some key takeouts regarding oil.

[00:03:03.050]
The first one is about the sensor location

[00:03:06.750]
now if you are uncertain about your sensor location

[00:03:10.110]
then your target location will also be uncertain.

[00:03:13.420]
Let’s break it down.

[00:03:15.790]
If your sensor location and uncertainty

[00:03:17.090]
say is plus or minus 10 meters

[00:03:20.470]
then you cannot expect a depth solution

[00:03:22.680]
that has a better uncertainty

[00:03:24.880]
than plus or minus 10 centimeters.

[00:03:30.950]
Radiant uncertainty, we’ve already covered this

[00:03:35.230]
and as previously mentioned gradients can be measured

[00:03:38.530]
or calculated from the total field.

[00:03:42.550]
For the structural index

[00:03:44.610]
it helps if you have a rough idea

[00:03:46.360]
of what you are looking for.

[00:03:49.090]
For example a structural index of two

[00:03:51.830]
would be applicable to pipes

[00:03:54.110]
and three would be applicable to barrels

[00:03:57.370]
or UXO objects.

[00:04:02.850]
Previous experiences have shown that if the structural index

[00:04:06.340]
is low then the depth estimation will be shallow.

[00:04:10.910]
If the structural index is high

[00:04:13.070]
then the depth estimation will be deep.

[00:04:16.070]
This is something to keep in mind.

[00:04:19.350]
Another interesting one that we should always think about

[00:04:22.330]
is the window size.

[00:04:25.280]
When you have closed overlapping targets

[00:04:27.520]
or complex targets that are close together

[00:04:30.530]
then this presents a problem

[00:04:32.560]
because the window size must be large enough

[00:04:34.830]
to include all the signal from the modeled object.

[00:04:39.860]
However, the larger the window size

[00:04:43.300]
the more likely that you will be modeling

[00:04:45.340]
more than one single object inside that window

[00:04:48.750]
which in turn may lead to us getting spurious results.

[00:04:53.170]
So these are some of the key takeouts on Euler.

[00:04:56.650]
So here is the total field of the MagDrone R3

[00:04:59.770]
flown over the test site.

[00:05:02.198]
We want to see if we can estimate the

[00:05:03.810]
depth of these barrels and pipes that are buried

[00:05:07.560]
at this test site.

[00:05:10.140]
from the total field we have calculated

[00:05:14.300]
our analytic signal which you can see on your screens.

[00:05:19.050]
And from the analytics signal we have picked some targets

[00:05:23.380]
whose depths we are going to be estimating using Euler.

[00:05:29.940]
Now once we have picked our targets

[00:05:32.960]
the next step is to calculate the footprint

[00:05:35.240]
of each individual anomaly.

[00:05:40.090]
What we do here is to calculate a window size

[00:05:42.840]
based on the inflection point around the anomalies.

[00:05:46.020]
And then display that window

[00:05:47.530]
with proportional sized symbols.

[00:05:51.920]
The goal is to have your window size

[00:05:54.280]
around your anomaly that is large enough

[00:05:56.410]
to capture the entire anomaly

[00:05:58.340]
and hopefully some of the background as well.

[00:06:02.060]
If you are not assembling the entire wavelength

[00:06:04.330]
you start having errors creep into your depth estimation.

[00:06:08.370]
Similarly if the window is too large

[00:06:10.430]
and you start to include significant effects

[00:06:13.670]
from multiple sources,

[00:06:16.530]
or if your window includes a large

[00:06:18.810]
number of now values which can happen for example

[00:06:23.440]
at the edge of our grid.

[00:06:27.920]
A good example would be if we had left

[00:06:30.180]
this as a target which is on the edge.

[00:06:33.460]
This is where you’re not going to capture the entire signal

[00:06:37.970]
that could result in errors in our depth estimation.

[00:06:42.580]
To estimate the sizes of anomaly of each target.

[00:06:46.430]
Click on UXO Marine Mag

[00:06:50.090]
and then scroll down to calculate anomaly sizes.

[00:06:56.420]
The grid to define the target sizes

[00:06:58.590]
is your analytic signal grid

[00:07:01.650]
and we have a target database which contains our target

[00:07:05.670]
and the group is called merged

[00:07:08.930]
and we are going to leave our mask channel as mask.

[00:07:14.290]
And we are going to call our output size channel

[00:07:18.418]
as size and then we click the okay button.

[00:07:23.370]
The size channel is then created.

[00:07:27.410]
The next step is to plot this size channel on top.

[00:07:33.750]
And it is plotted as a square with a half width.

[00:07:40.380]
Now something that we need to know here

[00:07:43.530]
is the scale of our map.

[00:07:46.430]
And this is a number

[00:07:47.330]
that we need to keep at the back of our minds

[00:07:49.810]
because we are going to use it in the next step.

[00:07:53.250]
To find the scale of your map if you click on map tools

[00:07:58.630]
then under map tools if you go to base map

[00:08:02.130]
and then click on draw base map

[00:08:06.040]
you will find the scale of your map.

[00:08:08.950]
In my case my map scale is 500.

[00:08:13.570]
I’m just going to keep this number at the back of my mind

[00:08:16.820]
because I’m going to use it in the next step.

[00:08:21.010]
And then to plot the size channel onto my map

[00:08:27.250]
I go to map tools and then from map tools I go to symbols,

[00:08:34.080]
from symbols I’m going to click under proportional size.

[00:08:40.860]
What I want to plot is my size channel

[00:08:44.200]
so I select size.

[00:08:47.690]
The zero base level we are going to leave this one as zero.

[00:08:52.250]
The scale factor, the way that you calculate it

[00:08:57.980]
you take the scale of your map

[00:09:01.160]
which in my case is 500

[00:09:03.770]
and you divide it by 1000 to get millimeters

[00:09:07.880]
which gives me 0.5.

[00:09:12.210]
Now my square is a half length.

[00:09:16.830]
So I need to divide this 0.5 by two

[00:09:21.280]
which gives me 0.25.

[00:09:25.950]
In my mask channel I’m going to leave it as mask

[00:09:30.700]
then come down here click on symbols

[00:09:34.460]
we’re going to choose the square symbol

[00:09:38.950]
the rest we are going to leave as default

[00:09:41.470]
and then you hit the okay button at the bottom,

[00:09:46.560]
and hit okay again.

[00:09:49.380]
Now we have plotted our squares

[00:09:52.660]
to estimate the sizes of anomalies for each target

[00:09:57.711]
let’s go back slightly.

[00:10:00.000]
I’ll deselect these squares.

[00:10:05.370]
One may ask why do pipes have multiple targets on them?

[00:10:11.696]
And for example a barrel that we have

[00:10:13.789]
for example at the bottom here

[00:10:15.690]
only has a single target.

[00:10:19.200]
This is because pipes are uniformly magnetized

[00:10:22.490]
along their length.

[00:10:24.720]
So the accuracy filter only allows solutions

[00:10:27.630]
at the pipe ends where there is sufficient

[00:10:30.580]
horizontal resolving power.

[00:10:34.290]
Let us bring back our square windows.

[00:10:39.550]
Remember earlier we mentioned that the goal with the windows

[00:10:43.290]
is to have them in such a way

[00:10:45.010]
that around the anomaly they are large enough

[00:10:48.700]
to capture the entire anomaly

[00:10:50.610]
and hopefully some of the background as well.

[00:10:54.637]
Let us pick one target of interest,

[00:11:00.520]
and zoom into it.

[00:11:04.490]
You will notice this square that I have here

[00:11:10.620]
it is small and does not capture the entire anomaly.

[00:11:16.950]
So there is a need for you to manipulate

[00:11:21.450]
and test the window sizes

[00:11:25.530]
in such a way that you capture it in its entirety.

[00:11:33.330]
So what i have done in the background

[00:11:36.540]
is to manipulate and do some trial and error

[00:11:41.410]
with the window sizes and edited this channel

[00:11:47.290]
to create a size edit channel.

[00:11:52.240]
This is now the channel that I am going to plot

[00:11:55.510]
onto my map.

[00:11:57.930]
To do this again I go to map tools,

[00:12:02.500]
scroll to symbols just like I did previously

[00:12:08.290]
select proportional size

[00:12:11.550]
but instead of the data channel being size

[00:12:15.180]
I’m going to choose size edit

[00:12:19.310]
and then I’m going to click okay.

[00:12:24.450]
The next thing that I’m going to do is to deselect

[00:12:28.230]
my initial size channel that I added first

[00:12:33.190]
and then you can see that

[00:12:36.030]
I am now capturing the signal

[00:12:39.580]
of the anomaly in its entirety.

[00:12:44.200]
The next step is for us

[00:12:47.810]
to calculate the depth using Euler deconvolution.

[00:12:53.000]
To do this you go to the menu UXO marine

[00:12:59.290]
and then scroll to calculate depth

[00:13:01.740]
and weights using Euler deconvolution.

[00:13:06.940]
Our target database is target, the group is merged

[00:13:13.300]
the mask channel I am going to change this

[00:13:18.270]
because i have two types of targets.

[00:13:20.700]
I am going to deal with pipes first to separate the two

[00:13:26.940]
and then the target type this is the structural index.

[00:13:31.220]
Since we are looking at pipes we are going to choose two

[00:13:37.280]
which applies to cylinders and pipes.

[00:13:42.680]
The target size channel

[00:13:45.030]
is going to be our size edit channel.

[00:13:49.150]
For the instrument height

[00:13:51.140]
we are going to leave it at zero.

[00:13:57.170]
By leaving the instrument height as zero

[00:14:00.310]
it means we are calculating

[00:14:02.580]
depth to be below the instrument

[00:14:05.760]
and one needs to go and manually subtract

[00:14:10.900]
the instrument height from the final depth estimates.

[00:14:18.210]
My analytical signal grid is given there

[00:14:23.800]
and the rest as well for the suit

[00:14:27.790]
and then I’m going to click the okay button.

[00:14:33.800]
Now what I have is a depth channel

[00:14:39.430]
an apparent weight channel

[00:14:42.380]
as a as well as an error channel.

[00:14:47.260]
I am going to plot this nicely together with my map

[00:14:52.140]
by going to window and say tile vertically.

[00:15:04.350]
And then I’m going to post

[00:15:07.020]
the depth channel onto my map

[00:15:09.980]
so we can see the estimated depth.

[00:15:14.790]
To post this depth channel onto my map

[00:15:18.550]
I go to map tools posting

[00:15:25.260]
the channel that I want to select

[00:15:26.930]
the channel that you want to post

[00:15:29.330]
which is mag_depth

[00:15:33.490]
and then I’m going to click the okay button.

[00:15:39.300]
Let’s just enlarge our map

[00:15:43.300]
so we can observe our targets.

[00:15:55.840]
So we can see

[00:15:58.900]
if we look at this pipe right here

[00:16:03.100]
it’s at a depth of between 2.8 and 2.4 meters.

[00:16:08.860]
The first one here is between 1.7 and 2.5.

[00:16:15.769]
What I’m going to do now is to pull in a document

[00:16:19.400]
with the information for our targets.

[00:16:24.730]
The first target is at a depth of 0.4 and 1 meter.

[00:16:31.350]
The second one is between one meter and two meters.

[00:16:37.920]
If we look at our first target it is at 1.7 and 2.5 meters.

[00:16:46.660]
If we subtract our flight height which is approximately

[00:16:50.627]
1.2 to 1.3 meters.

[00:16:54.960]
Then it means our target here is between 0.4

[00:17:00.660]
and 1.2 meters.

[00:17:05.210]
So this is acceptable and if you have a channel

[00:17:11.460]
that has your altitude that you would have calculated

[00:17:15.610]
you can then easily subtract it by using channel math here

[00:17:21.080]
if you go to database tools,

[00:17:24.130]
channel math you have a simple equation

[00:17:28.560]
for doing the calculation.

[00:17:32.110]
So basically that’s about it.

[00:17:35.410]
If you have any questions

[00:17:37.410]
please feel free to reach out to us via email

[00:17:40.580]
using [email protected]

[00:17:44.590]
thanks again for watching folks

[00:17:46.850]
have yourselves a wonderful day.

[00:00:00.217]
(ethereal music)

[00:00:10.600]
<v Victor>Thank you for dialing in today.</v>

[00:00:13.830]
I hope you have enjoyed Lyceum thus far.

[00:00:16.990]
My name is Victor,

[00:00:17.850]
and I’m a customer solution specialist at Seequent.

[00:00:21.480]
The title of my presentation is Euler

[00:00:24.177]
:a classic beat that still to the depths of geophysics.

[00:00:28.810]
I would like to acknowledge as pH engineering

[00:00:31.070]
we have shared their magnetic data acquired over a test site

[00:00:34.877]
at Bellow in Latvia,

[00:00:37.020]
using the synthesized MagDrone R3.

[00:00:42.647]
The test site consists of a number of buried ferrous

[00:00:45.380]
and non-ferrous targets.

[00:00:47.550]
Examples of which you can see in front of your screen

[00:00:51.360]
such as pipes and barrels.

[00:00:54.760]
The purpose of this presentation

[00:00:57.050]
is to show some tools that are available

[00:00:58.960]
in Oasis montaj that can be used

[00:01:01.240]
to estimate the depth of these objects.

[00:01:04.420]
Today we are going to focus on Euler deconvolution.

[00:01:09.780]
This is applicable in mapping objects such as utilities.

[00:01:14.080]
You can also use this method to estimate

[00:01:16.210]
the depth of unexploded ordinance.

[00:01:19.600]
We’ll be going through some workflows

[00:01:21.320]
for the estimation of depth for these types of objects.

[00:01:28.260]
Perhaps as a form of introduction

[00:01:31.030]
let us go over the Euler homogeneity equation

[00:01:35.460]
x, y as well as z is a location

[00:01:41.440]
or measurement point so to say.

[00:01:46.770]
X naught, y naught are know from our pre-target locations

[00:01:52.210]
and we are solving for z naught

[00:01:54.710]
which is the depth of the object.

[00:01:58.850]
The derivatives delta t over delta x,

[00:02:03.030]
delta t over delta y, and delta t over delta z

[00:02:07.660]
represent the gradients.

[00:02:11.090]
If you are using a total field magnetometer

[00:02:14.320]
you can derive these gradients.

[00:02:17.870]
If you are using a flux gate magnetometer

[00:02:20.610]
or multiple total field magnetometers in an array

[00:02:25.770]
one or all of these gradients can be measured directly.

[00:02:32.280]
And then t represents the measured field.

[00:02:37.610]
And b represents our background field

[00:02:41.630]
we call N our structural index

[00:02:44.370]
and it accounts for the rate of decrease

[00:02:46.380]
in the amplitude of the response

[00:02:48.050]
with distance from the source.

[00:02:50.830]
It ranges from 0 to three for potential field data.

[00:02:56.050]
All right, let’s go over some key takeouts regarding oil.

[00:03:03.050]
The first one is about the sensor location

[00:03:06.750]
now if you are uncertain about your sensor location

[00:03:10.110]
then your target location will also be uncertain.

[00:03:13.420]
Let’s break it down.

[00:03:15.790]
If your sensor location and uncertainty

[00:03:17.090]
say is plus or minus 10 meters

[00:03:20.470]
then you cannot expect a depth solution

[00:03:22.680]
that has a better uncertainty

[00:03:24.880]
than plus or minus 10 centimeters.

[00:03:30.950]
Radiant uncertainty, we’ve already covered this

[00:03:35.230]
and as previously mentioned gradients can be measured

[00:03:38.530]
or calculated from the total field.

[00:03:42.550]
For the structural index

[00:03:44.610]
it helps if you have a rough idea

[00:03:46.360]
of what you are looking for.

[00:03:49.090]
For example a structural index of two

[00:03:51.830]
would be applicable to pipes

[00:03:54.110]
and three would be applicable to barrels

[00:03:57.370]
or UXO objects.

[00:04:02.850]
Previous experiences have shown that if the structural index

[00:04:06.340]
is low then the depth estimation will be shallow.

[00:04:10.910]
If the structural index is high

[00:04:13.070]
then the depth estimation will be deep.

[00:04:16.070]
This is something to keep in mind.

[00:04:19.350]
Another interesting one that we should always think about

[00:04:22.330]
is the window size.

[00:04:25.280]
When you have closed overlapping targets

[00:04:27.520]
or complex targets that are close together

[00:04:30.530]
then this presents a problem

[00:04:32.560]
because the window size must be large enough

[00:04:34.830]
to include all the signal from the modeled object.

[00:04:39.860]
However, the larger the window size

[00:04:43.300]
the more likely that you will be modeling

[00:04:45.340]
more than one single object inside that window

[00:04:48.750]
which in turn may lead to us getting spurious results.

[00:04:53.170]
So these are some of the key takeouts on Euler.

[00:04:56.650]
So here is the total field of the MagDrone R3

[00:04:59.770]
flown over the test site.

[00:05:02.198]
We want to see if we can estimate the

[00:05:03.810]
depth of these barrels and pipes that are buried

[00:05:07.560]
at this test site.

[00:05:10.140]
from the total field we have calculated

[00:05:14.300]
our analytic signal which you can see on your screens.

[00:05:19.050]
And from the analytics signal we have picked some targets

[00:05:23.380]
whose depths we are going to be estimating using Euler.

[00:05:29.940]
Now once we have picked our targets

[00:05:32.960]
the next step is to calculate the footprint

[00:05:35.240]
of each individual anomaly.

[00:05:40.090]
What we do here is to calculate a window size

[00:05:42.840]
based on the inflection point around the anomalies.

[00:05:46.020]
And then display that window

[00:05:47.530]
with proportional sized symbols.

[00:05:51.920]
The goal is to have your window size

[00:05:54.280]
around your anomaly that is large enough

[00:05:56.410]
to capture the entire anomaly

[00:05:58.340]
and hopefully some of the background as well.

[00:06:02.060]
If you are not assembling the entire wavelength

[00:06:04.330]
you start having errors creep into your depth estimation.

[00:06:08.370]
Similarly if the window is too large

[00:06:10.430]
and you start to include significant effects

[00:06:13.670]
from multiple sources,

[00:06:16.530]
or if your window includes a large

[00:06:18.810]
number of now values which can happen for example

[00:06:23.440]
at the edge of our grid.

[00:06:27.920]
A good example would be if we had left

[00:06:30.180]
this as a target which is on the edge.

[00:06:33.460]
This is where you’re not going to capture the entire signal

[00:06:37.970]
that could result in errors in our depth estimation.

[00:06:42.580]
To estimate the sizes of anomaly of each target.

[00:06:46.430]
Click on UXO Marine Mag

[00:06:50.090]
and then scroll down to calculate anomaly sizes.

[00:06:56.420]
The grid to define the target sizes

[00:06:58.590]
is your analytic signal grid

[00:07:01.650]
and we have a target database which contains our target

[00:07:05.670]
and the group is called merged

[00:07:08.930]
and we are going to leave our mask channel as mask.

[00:07:14.290]
And we are going to call our output size channel

[00:07:18.418]
as size and then we click the okay button.

[00:07:23.370]
The size channel is then created.

[00:07:27.410]
The next step is to plot this size channel on top.

[00:07:33.750]
And it is plotted as a square with a half width.

[00:07:40.380]
Now something that we need to know here

[00:07:43.530]
is the scale of our map.

[00:07:46.430]
And this is a number

[00:07:47.330]
that we need to keep at the back of our minds

[00:07:49.810]
because we are going to use it in the next step.

[00:07:53.250]
To find the scale of your map if you click on map tools

[00:07:58.630]
then under map tools if you go to base map

[00:08:02.130]
and then click on draw base map

[00:08:06.040]
you will find the scale of your map.

[00:08:08.950]
In my case my map scale is 500.

[00:08:13.570]
I’m just going to keep this number at the back of my mind

[00:08:16.820]
because I’m going to use it in the next step.

[00:08:21.010]
And then to plot the size channel onto my map

[00:08:27.250]
I go to map tools and then from map tools I go to symbols,

[00:08:34.080]
from symbols I’m going to click under proportional size.

[00:08:40.860]
What I want to plot is my size channel

[00:08:44.200]
so I select size.

[00:08:47.690]
The zero base level we are going to leave this one as zero.

[00:08:52.250]
The scale factor, the way that you calculate it

[00:08:57.980]
you take the scale of your map

[00:09:01.160]
which in my case is 500

[00:09:03.770]
and you divide it by 1000 to get millimeters

[00:09:07.880]
which gives me 0.5.

[00:09:12.210]
Now my square is a half length.

[00:09:16.830]
So I need to divide this 0.5 by two

[00:09:21.280]
which gives me 0.25.

[00:09:25.950]
In my mask channel I’m going to leave it as mask

[00:09:30.700]
then come down here click on symbols

[00:09:34.460]
we’re going to choose the square symbol

[00:09:38.950]
the rest we are going to leave as default

[00:09:41.470]
and then you hit the okay button at the bottom,

[00:09:46.560]
and hit okay again.

[00:09:49.380]
Now we have plotted our squares

[00:09:52.660]
to estimate the sizes of anomalies for each target

[00:09:57.711]
let’s go back slightly.

[00:10:00.000]
I’ll deselect these squares.

[00:10:05.370]
One may ask why do pipes have multiple targets on them?

[00:10:11.696]
And for example a barrel that we have

[00:10:13.789]
for example at the bottom here

[00:10:15.690]
only has a single target.

[00:10:19.200]
This is because pipes are uniformly magnetized

[00:10:22.490]
along their length.

[00:10:24.720]
So the accuracy filter only allows solutions

[00:10:27.630]
at the pipe ends where there is sufficient

[00:10:30.580]
horizontal resolving power.

[00:10:34.290]
Let us bring back our square windows.

[00:10:39.550]
Remember earlier we mentioned that the goal with the windows

[00:10:43.290]
is to have them in such a way

[00:10:45.010]
that around the anomaly they are large enough

[00:10:48.700]
to capture the entire anomaly

[00:10:50.610]
and hopefully some of the background as well.

[00:10:54.637]
Let us pick one target of interest,

[00:11:00.520]
and zoom into it.

[00:11:04.490]
You will notice this square that I have here

[00:11:10.620]
it is small and does not capture the entire anomaly.

[00:11:16.950]
So there is a need for you to manipulate

[00:11:21.450]
and test the window sizes

[00:11:25.530]
in such a way that you capture it in its entirety.

[00:11:33.330]
So what i have done in the background

[00:11:36.540]
is to manipulate and do some trial and error

[00:11:41.410]
with the window sizes and edited this channel

[00:11:47.290]
to create a size edit channel.

[00:11:52.240]
This is now the channel that I am going to plot

[00:11:55.510]
onto my map.

[00:11:57.930]
To do this again I go to map tools,

[00:12:02.500]
scroll to symbols just like I did previously

[00:12:08.290]
select proportional size

[00:12:11.550]
but instead of the data channel being size

[00:12:15.180]
I’m going to choose size edit

[00:12:19.310]
and then I’m going to click okay.

[00:12:24.450]
The next thing that I’m going to do is to deselect

[00:12:28.230]
my initial size channel that I added first

[00:12:33.190]
and then you can see that

[00:12:36.030]
I am now capturing the signal

[00:12:39.580]
of the anomaly in its entirety.

[00:12:44.200]
The next step is for us

[00:12:47.810]
to calculate the depth using Euler deconvolution.

[00:12:53.000]
To do this you go to the menu UXO marine

[00:12:59.290]
and then scroll to calculate depth

[00:13:01.740]
and weights using Euler deconvolution.

[00:13:06.940]
Our target database is target, the group is merged

[00:13:13.300]
the mask channel I am going to change this

[00:13:18.270]
because i have two types of targets.

[00:13:20.700]
I am going to deal with pipes first to separate the two

[00:13:26.940]
and then the target type this is the structural index.

[00:13:31.220]
Since we are looking at pipes we are going to choose two

[00:13:37.280]
which applies to cylinders and pipes.

[00:13:42.680]
The target size channel

[00:13:45.030]
is going to be our size edit channel.

[00:13:49.150]
For the instrument height

[00:13:51.140]
we are going to leave it at zero.

[00:13:57.170]
By leaving the instrument height as zero

[00:14:00.310]
it means we are calculating

[00:14:02.580]
depth to be below the instrument

[00:14:05.760]
and one needs to go and manually subtract

[00:14:10.900]
the instrument height from the final depth estimates.

[00:14:18.210]
My analytical signal grid is given there

[00:14:23.800]
and the rest as well for the suit

[00:14:27.790]
and then I’m going to click the okay button.

[00:14:33.800]
Now what I have is a depth channel

[00:14:39.430]
an apparent weight channel

[00:14:42.380]
as a as well as an error channel.

[00:14:47.260]
I am going to plot this nicely together with my map

[00:14:52.140]
by going to window and say tile vertically.

[00:15:04.350]
And then I’m going to post

[00:15:07.020]
the depth channel onto my map

[00:15:09.980]
so we can see the estimated depth.

[00:15:14.790]
To post this depth channel onto my map

[00:15:18.550]
I go to map tools posting

[00:15:25.260]
the channel that I want to select

[00:15:26.930]
the channel that you want to post

[00:15:29.330]
which is mag_depth

[00:15:33.490]
and then I’m going to click the okay button.

[00:15:39.300]
Let’s just enlarge our map

[00:15:43.300]
so we can observe our targets.

[00:15:55.840]
So we can see

[00:15:58.900]
if we look at this pipe right here

[00:16:03.100]
it’s at a depth of between 2.8 and 2.4 meters.

[00:16:08.860]
The first one here is between 1.7 and 2.5.

[00:16:15.769]
What I’m going to do now is to pull in a document

[00:16:19.400]
with the information for our targets.

[00:16:24.730]
The first target is at a depth of 0.4 and 1 meter.

[00:16:31.350]
The second one is between one meter and two meters.

[00:16:37.920]
If we look at our first target it is at 1.7 and 2.5 meters.

[00:16:46.660]
If we subtract our flight height which is approximately

[00:16:50.627]
1.2 to 1.3 meters.

[00:16:54.960]
Then it means our target here is between 0.4

[00:17:00.660]
and 1.2 meters.

[00:17:05.210]
So this is acceptable and if you have a channel

[00:17:11.460]
that has your altitude that you would have calculated

[00:17:15.610]
you can then easily subtract it by using channel math here

[00:17:21.080]
if you go to database tools,

[00:17:24.130]
channel math you have a simple equation

[00:17:28.560]
for doing the calculation.

[00:17:32.110]
So basically that’s about it.

[00:17:35.410]
If you have any questions

[00:17:37.410]
please feel free to reach out to us via email

[00:17:40.580]
using [email protected]

[00:17:44.590]
[00:00:00.217]
(ethereal music)

[00:00:10.600]
<v Victor>Thank you for dialing in today.</v>

[00:00:13.830]
I hope you have enjoyed Lyceum thus far.

[00:00:16.990]
My name is Victor,

[00:00:17.850]
and I’m a customer solution specialist at Seequent.

[00:00:21.480]
The title of my presentation is Euler

[00:00:24.177]
:a classic beat that still to the depths of geophysics.

[00:00:28.810]
I would like to acknowledge as pH engineering

[00:00:31.070]
we have shared their magnetic data acquired over a test site

[00:00:34.877]
at Bellow in Latvia,

[00:00:37.020]
using the synthesized MagDrone R3.

[00:00:42.647]
The test site consists of a number of buried ferrous

[00:00:45.380]
and non-ferrous targets.

[00:00:47.550]
Examples of which you can see in front of your screen

[00:00:51.360]
such as pipes and barrels.

[00:00:54.760]
The purpose of this presentation

[00:00:57.050]
is to show some tools that are available

[00:00:58.960]
in Oasis montaj that can be used

[00:01:01.240]
to estimate the depth of these objects.

[00:01:04.420]
Today we are going to focus on Euler deconvolution.

[00:01:09.780]
This is applicable in mapping objects such as utilities.

[00:01:14.080]
You can also use this method to estimate

[00:01:16.210]
the depth of unexploded ordinance.

[00:01:19.600]
We’ll be going through some workflows

[00:01:21.320]
for the estimation of depth for these types of objects.

[00:01:28.260]
Perhaps as a form of introduction

[00:01:31.030]
let us go over the Euler homogeneity equation

[00:01:35.460]
x, y as well as z is a location

[00:01:41.440]
or measurement point so to say.

[00:01:46.770]
X naught, y naught are know from our pre-target locations

[00:01:52.210]
and we are solving for z naught

[00:01:54.710]
which is the depth of the object.

[00:01:58.850]
The derivatives delta t over delta x,

[00:02:03.030]
delta t over delta y, and delta t over delta z

[00:02:07.660]
represent the gradients.

[00:02:11.090]
If you are using a total field magnetometer

[00:02:14.320]
you can derive these gradients.

[00:02:17.870]
If you are using a flux gate magnetometer

[00:02:20.610]
or multiple total field magnetometers in an array

[00:02:25.770]
one or all of these gradients can be measured directly.

[00:02:32.280]
And then t represents the measured field.

[00:02:37.610]
And b represents our background field

[00:02:41.630]
we call N our structural index

[00:02:44.370]
and it accounts for the rate of decrease

[00:02:46.380]
in the amplitude of the response

[00:02:48.050]
with distance from the source.

[00:02:50.830]
It ranges from 0 to three for potential field data.

[00:02:56.050]
All right, let’s go over some key takeouts regarding oil.

[00:03:03.050]
The first one is about the sensor location

[00:03:06.750]
now if you are uncertain about your sensor location

[00:03:10.110]
then your target location will also be uncertain.

[00:03:13.420]
Let’s break it down.

[00:03:15.790]
If your sensor location and uncertainty

[00:03:17.090]
say is plus or minus 10 meters

[00:03:20.470]
then you cannot expect a depth solution

[00:03:22.680]
that has a better uncertainty

[00:03:24.880]
than plus or minus 10 centimeters.

[00:03:30.950]
Radiant uncertainty, we’ve already covered this

[00:03:35.230]
and as previously mentioned gradients can be measured

[00:03:38.530]
or calculated from the total field.

[00:03:42.550]
For the structural index

[00:03:44.610]
it helps if you have a rough idea

[00:03:46.360]
of what you are looking for.

[00:03:49.090]
For example a structural index of two

[00:03:51.830]
would be applicable to pipes

[00:03:54.110]
and three would be applicable to barrels

[00:03:57.370]
or UXO objects.

[00:04:02.850]
Previous experiences have shown that if the structural index

[00:04:06.340]
is low then the depth estimation will be shallow.

[00:04:10.910]
If the structural index is high

[00:04:13.070]
then the depth estimation will be deep.

[00:04:16.070]
This is something to keep in mind.

[00:04:19.350]
Another interesting one that we should always think about

[00:04:22.330]
is the window size.

[00:04:25.280]
When you have closed overlapping targets

[00:04:27.520]
or complex targets that are close together

[00:04:30.530]
then this presents a problem

[00:04:32.560]
because the window size must be large enough

[00:04:34.830]
to include all the signal from the modeled object.

[00:04:39.860]
However, the larger the window size

[00:04:43.300]
the more likely that you will be modeling

[00:04:45.340]
more than one single object inside that window

[00:04:48.750]
which in turn may lead to us getting spurious results.

[00:04:53.170]
So these are some of the key takeouts on Euler.

[00:04:56.650]
So here is the total field of the MagDrone R3

[00:04:59.770]
flown over the test site.

[00:05:02.198]
We want to see if we can estimate the

[00:05:03.810]
depth of these barrels and pipes that are buried

[00:05:07.560]
at this test site.

[00:05:10.140]
from the total field we have calculated

[00:05:14.300]
our analytic signal which you can see on your screens.

[00:05:19.050]
And from the analytics signal we have picked some targets

[00:05:23.380]
whose depths we are going to be estimating using Euler.

[00:05:29.940]
Now once we have picked our targets

[00:05:32.960]
the next step is to calculate the footprint

[00:05:35.240]
of each individual anomaly.

[00:05:40.090]
What we do here is to calculate a window size

[00:05:42.840]
based on the inflection point around the anomalies.

[00:05:46.020]
And then display that window

[00:05:47.530]
with proportional sized symbols.

[00:05:51.920]
The goal is to have your window size

[00:05:54.280]
around your anomaly that is large enough

[00:05:56.410]
to capture the entire anomaly

[00:05:58.340]
and hopefully some of the background as well.

[00:06:02.060]
If you are not assembling the entire wavelength

[00:06:04.330]
you start having errors creep into your depth estimation.

[00:06:08.370]
Similarly if the window is too large

[00:06:10.430]
and you start to include significant effects

[00:06:13.670]
from multiple sources,

[00:06:16.530]
or if your window includes a large

[00:06:18.810]
number of now values which can happen for example

[00:06:23.440]
at the edge of our grid.

[00:06:27.920]
A good example would be if we had left

[00:06:30.180]
this as a target which is on the edge.

[00:06:33.460]
This is where you’re not going to capture the entire signal

[00:06:37.970]
that could result in errors in our depth estimation.

[00:06:42.580]
To estimate the sizes of anomaly of each target.

[00:06:46.430]
Click on UXO Marine Mag

[00:06:50.090]
and then scroll down to calculate anomaly sizes.

[00:06:56.420]
The grid to define the target sizes

[00:06:58.590]
is your analytic signal grid

[00:07:01.650]
and we have a target database which contains our target

[00:07:05.670]
and the group is called merged

[00:07:08.930]
and we are going to leave our mask channel as mask.

[00:07:14.290]
And we are going to call our output size channel

[00:07:18.418]
as size and then we click the okay button.

[00:07:23.370]
The size channel is then created.

[00:07:27.410]
The next step is to plot this size channel on top.

[00:07:33.750]
And it is plotted as a square with a half width.

[00:07:40.380]
Now something that we need to know here

[00:07:43.530]
is the scale of our map.

[00:07:46.430]
And this is a number

[00:07:47.330]
that we need to keep at the back of our minds

[00:07:49.810]
because we are going to use it in the next step.

[00:07:53.250]
To find the scale of your map if you click on map tools

[00:07:58.630]
then under map tools if you go to base map

[00:08:02.130]
and then click on draw base map

[00:08:06.040]
you will find the scale of your map.

[00:08:08.950]
In my case my map scale is 500.

[00:08:13.570]
I’m just going to keep this number at the back of my mind

[00:08:16.820]
because I’m going to use it in the next step.

[00:08:21.010]
And then to plot the size channel onto my map

[00:08:27.250]
I go to map tools and then from map tools I go to symbols,

[00:08:34.080]
from symbols I’m going to click under proportional size.

[00:08:40.860]
What I want to plot is my size channel

[00:08:44.200]
so I select size.

[00:08:47.690]
The zero base level we are going to leave this one as zero.

[00:08:52.250]
The scale factor, the way that you calculate it

[00:08:57.980]
you take the scale of your map

[00:09:01.160]
which in my case is 500

[00:09:03.770]
and you divide it by 1000 to get millimeters

[00:09:07.880]
which gives me 0.5.

[00:09:12.210]
Now my square is a half length.

[00:09:16.830]
So I need to divide this 0.5 by two

[00:09:21.280]
which gives me 0.25.

[00:09:25.950]
In my mask channel I’m going to leave it as mask

[00:09:30.700]
then come down here click on symbols

[00:09:34.460]
we’re going to choose the square symbol

[00:09:38.950]
the rest we are going to leave as default

[00:09:41.470]
and then you hit the okay button at the bottom,

[00:09:46.560]
and hit okay again.

[00:09:49.380]
Now we have plotted our squares

[00:09:52.660]
to estimate the sizes of anomalies for each target

[00:09:57.711]
let’s go back slightly.

[00:10:00.000]
I’ll deselect these squares.

[00:10:05.370]
One may ask why do pipes have multiple targets on them?

[00:10:11.696]
And for example a barrel that we have

[00:10:13.789]
for example at the bottom here

[00:10:15.690]
only has a single target.

[00:10:19.200]
This is because pipes are uniformly magnetized

[00:10:22.490]
along their length.

[00:10:24.720]
So the accuracy filter only allows solutions

[00:10:27.630]
at the pipe ends where there is sufficient

[00:10:30.580]
horizontal resolving power.

[00:10:34.290]
Let us bring back our square windows.

[00:10:39.550]
Remember earlier we mentioned that the goal with the windows

[00:10:43.290]
is to have them in such a way

[00:10:45.010]
that around the anomaly they are large enough

[00:10:48.700]
to capture the entire anomaly

[00:10:50.610]
and hopefully some of the background as well.

[00:10:54.637]
Let us pick one target of interest,

[00:11:00.520]
and zoom into it.

[00:11:04.490]
You will notice this square that I have here

[00:11:10.620]
it is small and does not capture the entire anomaly.

[00:11:16.950]
So there is a need for you to manipulate

[00:11:21.450]
and test the window sizes

[00:11:25.530]
in such a way that you capture it in its entirety.

[00:11:33.330]
So what i have done in the background

[00:11:36.540]
is to manipulate and do some trial and error

[00:11:41.410]
with the window sizes and edited this channel

[00:11:47.290]
to create a size edit channel.

[00:11:52.240]
This is now the channel that I am going to plot

[00:11:55.510]
onto my map.

[00:11:57.930]
To do this again I go to map tools,

[00:12:02.500]
scroll to symbols just like I did previously

[00:12:08.290]
select proportional size

[00:12:11.550]
but instead of the data channel being size

[00:12:15.180]
I’m going to choose size edit

[00:12:19.310]
and then I’m going to click okay.

[00:12:24.450]
The next thing that I’m going to do is to deselect

[00:12:28.230]
my initial size channel that I added first

[00:12:33.190]
and then you can see that

[00:12:36.030]
I am now capturing the signal

[00:12:39.580]
of the anomaly in its entirety.

[00:12:44.200]
The next step is for us

[00:12:47.810]
to calculate the depth using Euler deconvolution.

[00:12:53.000]
To do this you go to the menu UXO marine

[00:12:59.290]
and then scroll to calculate depth

[00:13:01.740]
and weights using Euler deconvolution.

[00:13:06.940]
Our target database is target, the group is merged

[00:13:13.300]
the mask channel I am going to change this

[00:13:18.270]
because i have two types of targets.

[00:13:20.700]
I am going to deal with pipes first to separate the two

[00:13:26.940]
and then the target type this is the structural index.

[00:13:31.220]
Since we are looking at pipes we are going to choose two

[00:13:37.280]
which applies to cylinders and pipes.

[00:13:42.680]
The target size channel

[00:13:45.030]
is going to be our size edit channel.

[00:13:49.150]
For the instrument height

[00:13:51.140]
we are going to leave it at zero.

[00:13:57.170]
By leaving the instrument height as zero

[00:14:00.310]
it means we are calculating

[00:14:02.580]
depth to be below the instrument

[00:14:05.760]
and one needs to go and manually subtract

[00:14:10.900]
the instrument height from the final depth estimates.

[00:14:18.210]
My analytical signal grid is given there

[00:14:23.800]
and the rest as well for the suit

[00:14:27.790]
and then I’m going to click the okay button.

[00:14:33.800]
Now what I have is a depth channel

[00:14:39.430]
an apparent weight channel

[00:14:42.380]
as a as well as an error channel.

[00:14:47.260]
I am going to plot this nicely together with my map

[00:14:52.140]
by going to window and say tile vertically.

[00:15:04.350]
And then I’m going to post

[00:15:07.020]
the depth channel onto my map

[00:15:09.980]
so we can see the estimated depth.

[00:15:14.790]
To post this depth channel onto my map

[00:15:18.550]
I go to map tools posting

[00:15:25.260]
the channel that I want to select

[00:15:26.930]
the channel that you want to post

[00:15:29.330]
which is mag_depth

[00:15:33.490]
and then I’m going to click the okay button.

[00:15:39.300]
Let’s just enlarge our map

[00:15:43.300]
so we can observe our targets.

[00:15:55.840]
So we can see

[00:15:58.900]
if we look at this pipe right here

[00:16:03.100]
it’s at a depth of between 2.8 and 2.4 meters.

[00:16:08.860]
The first one here is between 1.7 and 2.5.

[00:16:15.769]
What I’m going to do now is to pull in a document

[00:16:19.400]
with the information for our targets.

[00:16:24.730]
The first target is at a depth of 0.4 and 1 meter.

[00:16:31.350]
The second one is between one meter and two meters.

[00:16:37.920]
If we look at our first target it is at 1.7 and 2.5 meters.

[00:16:46.660]
If we subtract our flight height which is approximately

[00:16:50.627]
1.2 to 1.3 meters.

[00:16:54.960]
Then it means our target here is between 0.4

[00:17:00.660]
and 1.2 meters.

[00:17:05.210]
So this is acceptable and if you have a channel

[00:17:11.460]
that has your altitude that you would have calculated

[00:17:15.610]
you can then easily subtract it by using channel math here

[00:17:21.080]
if you go to database tools,

[00:17:24.130]
channel math you have a simple equation

[00:17:28.560]
for doing the calculation.

[00:17:32.110]
So basically that’s about it.

[00:17:35.410]
If you have any questions

[00:17:37.410]
please feel free to reach out to us via email

[00:17:40.580]
using [email protected]

[00:17:44.590]
thanks again for watching folks

[00:17:46.850]
have yourselves a wonderful day.

[00:17:49.698]
(ethereal music)

thanks again for watching folks

[00:17:46.850]
have yourselves a wonderful day.

[00:17:49.698]
(ethereal music)

[00:17:49.698]
(ethereal music)