Zigzag design - Insufficient sampler length

[Jérôme Couvat]

Dear all,

I am trying to design an Equal Spaced Zigzag design in a survey area divided in 2 strata. My survey area is a watermass around the island of Martinique.

Stratum 1 accounts for 70% of the study area and stratum 2 for 30%.

Because of financial limitations, I know that my total distance on effort for surveying this area is 776km, which I intend to divide proportionnally to the size of each stratum:

 - stratum 1 : 543km

 - stratum 2 : 233km

When I try to design my survey with such effort allocation, I get the following message "Insufficient sampler length to calculate a zigzag spacing" for stratum 2.

However, if I design a survey on stratum 2 only on a separate project, with the same effort allocation (233km), Distance has no problem designing it.

Any idea what I might be doing wrong?

Thanks a lot!

Jérôme Couvat

[Eric Rexstad]

Jérôme

I do not have the definitive answer to your question.  There are a number of factors that influence the design engine's ability to place effort within the study area or stratum; such as the shape of the study area, the design angle specified.  It may not be the case that you are doing anything "wrong".  Possibly you might need to experiment a bit with the design angles or adjust the stratum-specific line lengths (juggle the values a bit: 530/246, 520/256) and see if you have success.

If you wish to send your shape file along, I could spend a bit of time investigating your situation.

--
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/9dd6cbfe-bd94-466c-a3bb-0ecc8a25f884n%40googlegroups.com.

-- 
Eric Rexstad
Centre for Ecological and Environmental Modelling
University of St Andrews
St Andrews is a charity registered in Scotland SC013532

[Jérôme Couvat]

---------- Forwarded message ---------
De : Eric Rexstad <er...@st-andrews.ac.uk>
Date: lun. 14 déc. 2020 à 08:52
Subject: Re: [distance-sampling] Zigzag design - Insufficient sampler length
To: Jérôme Couvat <jeco...@gmail.com>

Happy to help; I will log the issue regarding the failure of the design engine to cope with your situation.  Thanks for forwarding information to the list.

On 14/12/2020 12:48, Jérôme Couvat wrote:

Thanks Eric. I was thinking of going around the problem as well so I am fine with that solution. Thank you very much for the time you spent on this. I will copy your response to the discussion to get the information to everyone. Thanks again.

Jérôme

Le sam. 12 déc. 2020 à 07:45, Eric Rexstad <er...@st-andrews.ac.uk> a écrit :

Jérôme

I have not done anything elegant, more the "brute force" approach.  I noticed in the project you sent, for some reason, the design engine had not calculated a spacing for the second stratum:

I don't know what that was not calculated, but instead of trying to solve that problem, I decided to go around the problem.  Instead of specifying line lengths, the design engine alternatively allows the user to specify the spacing of the waypoints, then computes transect lengths based upon the desired spacing.  I tried this approach:

I fiddled a bit with spacing values for Stratum 2 until I arrived at an effort allocation close to your desired 70/30 ratio.

This set of input seemed to please the design engine, as it carried on and generated a coverage for this survey design and subsequently generated a single realisation of the design to show this survey.

I realise it is not an elegant solution; I can't explain why your approach did not work.  But the work-around might get you the result you need to design your study.

If you are content with this approach, I would appreciate if you would write back to the list with this solution.  That way, should anyone else have this difficulty, they can benefit from the work-around.  Let me know if this works for you.

On 11/12/2020 20:11, Jérôme Couvat wrote:

Thank you Eric for your quick response. I did try to adjust design angles or effort length. Angles did not change a thing, although my stratum is far from being a perfect square. Line length was never a success either, even if I dramatically increase it to unrealistically-long distances... That worries me.

I've attached the files if you want to have a look. "Martinique_Project" includes the whole study area. "Stratum2" includes the successful design when I only consider the second stratum.

Thank you very much for your help.

Best,

Jérôme

[Jérôme Couvat]

Hi Eric,

I just tried that and it does not work for me. No matter the spacing I enter (1km, 100km, 10,000km), I still get 99% coverage for that stratum. Something is wrong somewhere, but I don't know where. Especially if it works fine for you... Is it worth uninstalling and re-installing Distance?

    

Jérôme

[Eric Rexstad]

Jérôme

No harm in reinstalling Distance; I don't expect that to have an effect, but no harm done.  Would you like me to send your project (upon which I experimented) back to you?

Sorry we are trying to resolve this problem on a day when Google is experiencing difficulties; messing with our ability to email each other.

To view this discussion on the web visit https://groups.google.com/d/msgid/distance-sampling/CA%2BWvfQ%2BCCWsP-puBQX1jeHf2X%2Bbgpj4rbfWNYEXL12gF8xDewA%40mail.gmail.com.

[Jérôme Couvat]

Yes please. Hopefully I will find something by comparing the two versions. If not, well I don't know...