Honeyeater migration A "me too" moment - more Yellow-faced Honeyeaters- ? cold cooma weather

[Philip Veerman <>]

Thanks for doing that. Yes I suggested “It may not even show well in a graph of that scale.” Nice to know. In this case the change is indeed trivial compared to the purpose of the graph. .

 

From: Martin Butterfield [
Sent: Saturday, 23 March, 2019 5:09 PM
To: Philip Veerman
Cc: canberrabirds
Subject: Re: [canberrabirds] Honeyeater migration A "me too" moment - more Yellow-faced Honeyeaters- ? cold cooma weather

 

In some cases - if considering a single year, and particularly if the week following Christmas is involved- adjusting for observer effort can make a difference.  That is why the A statistics is used.  

 

However I had found in the past that if one is looking across years - ie all records in week 'n' regardless of year - then the differences are trivial so its a waste of time going to the extra fiddling about as the shape of the chart is almost identical.  

 

As the question has been raised I have investigated the impact of adjusting for number of observer weeks. The observer effort is almost identical in each of the weeks shown - within 2% of the average.  The graphs are in effect identical in shape. (I have omitted  furniture such as titles and identifiers as they aren't necessary to demonstrate this point.)

 

Martin Butterfield

http://franmart.blogspot.com.au/

 

 

On Sat, 23 Mar 2019 at 16:23, Philip Veerman <> wrote:

Oh and one other thing, is that simply adding the totals of a species counted in the GBS by week, does not take into account the wide variation in the number of observer weeks by week. So giving varying sample sizes of the counts. These data are shown in Figure 5 “observer weeks by month” and  Figure 6 “observer weeks by week” of The GBS Report (page 26). Yes the differences for a small number of weeks over such a long period is tolerably small, given so many other awkward features. It may not even show well in a graph of that scale. Of course the difference is greatest during December and such a graph would be inappropriate then but it also is a small issue at the variable set of weeks that include Easter.

 

Philip

 

From: Philip Veerman [
Sent: Saturday, 23 March, 2019 10:07 AM
To: 'COG List'
Subject: RE: [canberrabirds] Honeyeater migration A "me too" moment - more Yellow-faced Honeyeaters- ? cold cooma weather

 

Yes a good graph but not notably unexpected. I also suggest that it is entirely consistent with what is showing this year. Historically, beginning the progress towards the peak is occurs this week. That is what is happening now. A few mentions of relatively small flocks on an email chat line does not constitute high levels so far. There has been little if anything more than the usual start to this process so far this year.

 

Also be aware that in the early years there was some clearly wonky counts of migrating YfH in the GBS. At that time, processes were not so well explained as in later versions of the chart or as I started doing the “Annual GBS promotional spectacular”. These were obtained as people clearly tallied counts over a period, rather than simultaneous. I recall one individual count of 10060. This is clearly not right. As I went through the process of compiling all those years of GBS results, as I recall I substituted that 10060 for 500. That is one I still remember now. Of course that was not right either but at least it is close to credible for an exceptionally large set of flocks that could have been in the area at one time. No way to know how many similar errors were made of this often hard to count phenomenon. I did not tell that story in The GBS Report but there is a strong hint of the problem of counting them.

 

I can’t guess what has happened since then (2002).