Verifying data quality and troubleshooting echosounders

1. How to verify data quality after deploying sensors.

Step 1: Sensor Placement and Installation

Ensure that sensors are placed according to EN Installation on Cage | STEP 5: Cage Duo Sensors

Step 2: Initial Data Verification

Check of the echograms for obvious problems with sensor positioning.

Step 3: Meal time data verification.

Verify that fish behavior is consistent with what you expect during mealtime.
If possible verify with cameras.

2. Troubleshooting echograms

Fish is outside of the measurement volume, that is below 25 meters or along the edges.

Empty Cage / No relevant signals leads to amplification of noise.

Fish taken out of cage, while equipment still operative. Inspector scales echograms by looking at the typical echostrenghts in the last 24 hours. Consequently, right after the cage has been emptied, the echogram is black as expected.

However, the day after, the echogram in the bottom sensor looking up here is very noisy. This is because, the lack of physical objects in the water column leads to very little echo and a normalization scale which will enhance electrical + acoustic noise. It is likely that this will manifest itself in the upward looking beam. Simply because the coverage volume is smaller (due to surface cropping ~25m) and hence more likely to be empty than the top sensor looking down (~50m).

Sensor is up from the water

Echosunders entangled with equipment

In this instance the echosounder was entangled with a lasernode.

Echograms are not showing small fish.

Image1: One color on the echogram is not visible on the graph.

Image2: With google chrome dark mode, you can see an additional color nuance in grey, which is invisible otherwise.

Echograms show a sawtooth stripe pattern.

27.Sept.23 - Western norway.

Symptoms: Echograms seens as ‘weird’ showing apparently artificial data on fish behavior on both Inspector and FA3(images).

Hypothesis: Seems like it is missing data from the accelerometer for some reason. Could be that the MCU is not getting the proper responses from the sensor.

Solution: Power cycling the echosounder.

Many small gaps in the echograms

27.Jan.24 - Troms Fylke, Northern norway
Severe storm at the site. Data is filtered out if sensors tilt a lot.

Tuna entagled with echosounder cable

2022-09-22 - Western norway.

Symptoms: Surface detection fails on echograms and then data looks bad on Inspector, affecting the deep sensor(up-facing).

Cause: There was a predator(yellowfin tuna) inside the cage that got its tail tangled with our cable and eventually died.