Information Geometrics preliminary MagArrow and MagEx data processing program download

Hello Dears,

I'm new to Geometrics products and interested in MagArrow. I have some questions please. As a geophysicist, I examined the attached data and applied heading compensation to the measured data. My questions are after applying the 4th difference filter (the last panel in the attached figure) which is very important for measuring data noise especially comTing from the drone. I found the range to be very high and exceeding +/- 0.1, even after applying low pass filter to reduces the sampling rate of data from 20 Hz to 10 Hz. I think one of the biggest advantages of the MagArrow is that the sensor is suspended 3 meters below the drone to cancel out the effect of the noise coming from the drone, so I need to explain that.

Secondly, in the attached file of the survey, the range of the total magnetic field data is about 43,000 nanoTesla, while when calculating the IGRF for that region based on longitude and latitude, I found that the range is about 49,000 nT, which is very far from the measured data? Could you please explain to me the answers to these questions because our goal is to explore minerals and not UXO.

Kind Regards,

Ahmed

@ahmed_ramadan_geo 

Hi Ahmed,

    The remaining noise after compensation is likely from 3 main sources:

1. The intrinsic sensor noise. At 10Hz, 5pT/rt-Hz rms sensor noise is equivalent to about 0.08 nT peak-to-peak noise.

2. Remaining drone noise. Our EM customers use 5m suspension cables, instead of 3m, to further reduce the drone interference. But this depends on your drone.

3. Compensation noise. The compass reading is very noisy. Part of it will enter the data after compensation.

    If you want to further reduce the noise, you can apply LPF at lower frequencies. For mineral explorations, I don't think you need 10Hz data.

    Regarding your other question, the coordinates in the sample file is NOT at the real survey site. We intentionally shifted the coordinates for privacy reasons. MagArrow is based on the scalar sensor technology. Its readings are absolute, not relative.

Dear Rui,

Thank you for your answers.

Bests,

Ahmed

Hello Dear Rui,

I have questions regarding MagArrow Magnetic data compensation, if you can answer please,

1- Does the heading error compensation that applied by surveying data processing program remove the heading effect due to direction only or calibrates the maneuver effect due to the swinging of the sensor as well?

2- Regarding the heading compensation flight, I am confused about the best way to do this. In the manual, it is mentioned that we can apply a cloverleaf pattern (this is usually used in airborne survey) or a survey pattern or we can do a 360 degree turn at a certain point. However, it is written that we can use the calibration data measured in the turn, can I use the data from the turn to apply the heading error correction separately for each flight instead of doing a heading flight. Regarding the cloverleaf pattern application, is the data also processed in the same way using surveying data processing program?

3-What is the importance of the base station magnetometer data in this correction as mentioned in the manual, I think we can apply it without base magnetometer?

4- Finally, is there a specific flight that can be done to make  maneuver correction (FOM) (roll, pitch and yaw) that we apply in conventional magnetic airborne survey? or We just resort to the LP filter.

Sorry for the long post

Kind Regards,

Ahmed  

@ahmed_ramadan_geo 

Regarding your questions:

1. It removes the maneuver effect due to the swing of the sensor as well.

2. There are really NO best way to do the compensation flight. There are two main requirements for a compensation flight: 1. Include as much roll/pitch/yaw motions as possible. 2. Avoid magnetic gradient. If you can fly the drone very high (>100m) over a low gradient area, a cloverleaf pattern is probably the best (as in traditional airborne surveys in which you can fly high to avoid the local gradient.). But this pattern is hard to fly, especially over a small area. 360-degree turn is simple and almost guaranteed to have no magnetic gradient in the compensation data since it is over the same location. If the MagArrow doesn't have much pitch and roll motion, 360-deg turn is sufficient. But it doesn't compensate well for pitch and roll maneuver noise (no much pitch/roll calibration). A survey pattern is between the two patterns described above. You may be able to use it, instead of doing a calibration flight. You can use the calibration file generation module to process the cloverleaf pattern.

3. Since the calibration flight is typically short, a base-station may not be necessary. Diurnal field change is usually not significant in a minute.

4. As I mentioned in 1., maneuver correction is also applied here. You can also use LP filter but make sure NOT to apply the heading compensation after the LP filter.

@rzhang 

Thanks for your answer, why don't apply heading compensation after LPF? 

What about using the sequence below as processing flow chart:

1- Raw data of 20 hz
3- Lp filter 40 fidicules (2seconds) >>> Maneuver correction
4- Remove base mag
5- Heading correction compensation using (survey data processing of MagArrow)
6- Lag Correction
7- Igrf Removal
8- Leveling and Microleveling

kind regards,

Ahmed 

Hi Ahmed,

    Depending on the corner frequency of the LPF, some heading related noise may be removed already by the LPF. The heading correction (the full maneuver compensation, not line-leveling) will then add the noise back to your data.

    In your flow chart, I would do heading correction before LPF. In addition, I would use raw data of 10Hz instead, because at certain parts of world, the AC power is at 50Hz. There will be 50Hz alias into 20Hz raw data.

@rzhang Thank you very much for your helpful answer. Could you please tell me how to reduce the data sampling rate from 20Hz to 10Hz using low pass filter, what is the exact value to use in terms of fiducials?

@ahmed_ramadan_geo The raw MagArrow data is at 1000Hz. When you export the raw data to .csv files using the Survey Manager, you can choose a different sample rate. You don't need the data processing software to decimate the 20Hz data to 10Hz.