Aerial surveys produces peak in distance 0

[Laura]

Hi all,

I am working with aerial distance sampling data. We conducted line transects from a helicopter flying at an altitude of 60 metres. For all observations directly beneath the helicopter, it was difficult to determine the exact distance, so all detections within 0–60 metres were recorded as distance = 0.

This creates a problem, as it produces an artificial peak at distance = 0, which in turn affects the fit of the detection probability function. One possible solution is to bin the data, but since all distances beyond 60 metres are recorded exactly, I would prefer to avoid that if possible.

Is there an alternative approach to deal with this issue, other than binning?

Thank you beforehand!

[Eric Rexstad]

Laura

Thanks for your question. It is common to have obstruction vision or, difficulty in measuring distance with aerial surveys. Clearly, the data as described will not produce reliable estimates of animal abundance.

One remedy (not without its difficulties) for this problem is to left-truncate the data. Field issues of aerial surveys are described in section 7.6 of Buckland et al. (2001). Section 4.3.2 of the same text describes two approaches to implementing left truncation to data collected in the field. Finally, examples implementing the two left truncation approaches are demonstrated in section 8.4.3 and Figure 8.8.

One challenge is to decide the left truncation distance. That may be a partially subjective decision, so best to be able to justify the choice of that distance. From your brief description, it seems you have no other choice but to truncate at 60m. Your next task is to decide which of the truncation implementation methods of Section 4.3.2 to implement. Comparison of estimates produced by the implementation methods for some duck nest data is shown in Table 8.6. However, I don't advocate performing both analyses and picking the result you like best. Think carefully about what the two approaches.

---

From: distance...@googlegroups.com <distance...@googlegroups.com> on behalf of Laura <laurabarbe...@gmail.com>
Sent: 17 September 2025 12:38
To: distance-sampling <distance...@googlegroups.com>
Subject: [distance-sampling] Aerial surveys produces peak in distance 0

 

--
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 visit https://groups.google.com/d/msgid/distance-sampling/cd64278d-fa5b-45c9-9017-d383336f82f4n%40googlegroups.com.

[Tiago Marques]

Hi Laura, 

A possible alternative approach to the analysis that Eric is suggesting, would be to set up a bespoke likelihood that uses binned data on the first 60 metres and exact data for detections further away from the line.

If my memory is not tricking me I might have done something like this a few years agi, but can't remember under what contrxt/project or where the corresponding code might be.

The downsides of this suggestion is that I don't remember anyone publishing an example of this working, but it should work. And, unfortunately, this is not available as an option in existing software AFAIK. Therefore, going that way  might require some coding expertise in creating your own bespoke likelihood, maximising it and harvesting the results to get density estimates and precision measures, a journey  which you might not want to embark on. This would possibly be an interesting MSc project for a student in an Ecological Statistics program, or for a chapter in a similarly veined methods oriented PhD, if there's one out there, or his/her supervisor, reading this list...

Cheers,

Tiago

Enviado de Outlook para Android

---

From: 'Eric Rexstad' via distance-sampling <distance...@googlegroups.com>
Sent: Wednesday, September 17, 2025 4:39:59 PM
To: Laura <laurabarbe...@gmail.com>; distance-sampling <distance...@googlegroups.com>
Subject: Re: [distance-sampling] Aerial surveys produces peak in distance 0

 

To view this discussion visit https://groups.google.com/d/msgid/distance-sampling/AM9PR06MB81460B8AB6CE52490925B8FBEA17A%40AM9PR06MB8146.eurprd06.prod.outlook.com.