Effective angle camera-trap distance sampling

[PABLO PALENCIA MAYORDOMO]

Dear everyone,

I'm currently working on simulations and distance sampling with camera-traps (CTDS). I'm thinking about the effect of the effective detection angle (EDA).

The expected distribution of angles within the FOV is uniform. However, some bias can be introduced because of the trigger speed or the sensitivity of the sensor at the periphery of its horizontal range.

One example:

Here I plot the detection zone of one camera-trap (located at coordinates -334, -334). Red dotted vertical line represents the center of the FOV of this camera (55 degrees). Blue points (15º < detection angle < 27.5º) and orange triangles (detection angle < 15º) represent the position of all animals on the first detection.

Here, I plot the histogram of detection angles (X-axis expressed in degrees). I considered the center of the FOV as 0 degrees, and I assumed symmetric detection to both sides (for that, the maximum detection angle observed was 27.5º)

As expected, most of the detection angles were 25 (approx.) which means fast activations by the camera. However, in some observations, detection angles were closer to 0 (center of the FOV) because of slow trigger speed and/or fast movements by the animals.

At this respect, when estimating density with CTDS, do you think that we have to include the angle described by the manufacture (maximum, here 55), or to estimate the effective detection angle?

In my opinion, estimating the EDA should be necessary to avoid underestimations of density when applying CTDS, especially when using cameras with slow trigger speed.

Following my example, I fitted detection function to my data:

(e.g.: ds(df_CTDS_ang_point$angle, transect = "line", key="hn", adjustment = NULL,
              truncation = 28))

However, I am not satisfied with the result... As the maximum observed frequency was at maximum angles, I'm not sure if it is possible to estimate EDA following classical line-transect distance sampling theory.

Thank you very much

Best

Pablo

[Stephen Buckland]

Pablo, yes, it is as well to estimate the angle.  The maximum angles you record are informative, but angles close to zero are not.  Animals are not assumed to be detected with certainty – the further the animal from the camera, the less likely (on average) it is to be detected.  So a more distant animal may well reach the centreline before it is detected.  In any case, animals might approach the camera from an angle that means they do not intersect the lines of your cone before detection.  In your diagram, only the shorter distances are useful for estimating the viewing angle, because probability of detection of an animal in the sector is close to one for those cases.  Or you could trigger the camera by moving in front of (and close to) the camera and recording the maximum angles at which it triggers.

 

Steve Buckland

--
You received this message because you are subscribed to the Google Groups "distance-sampling" group.
To unsubscribe from this group and stop receiving emails from it, send an email to distance-sampl...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/distance-sampling/ae82d46e-8741-4149-b5e6-611c9ad29405n%40googlegroups.com.

[PABLO PALENCIA MAYORDOMO]

Thank you very much for your reply Steve!

Makes sense to estimate effective angle just considering closer animals in which -radial- prob of detection is 1. 

I got it!

Thanks again!

Pablo

Obtener Outlook para Android

---

From: Stephen Buckland <st...@st-andrews.ac.uk>
Sent: Friday, February 5, 2021 9:17:23 PM
To: PABLO PALENCIA MAYORDOMO <palencia...@gmail.com>; distance-sampling <distance...@googlegroups.com>
Subject: RE: {Suspected Spam} [distance-sampling] Effective angle camera-trap distance sampling