I just compared Mike Patterson's compilation of Snowy Owl numbers to
the Oceanic Ni?o Index at
http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ensoyears.shtml.
While the analysis is eyeball only, a pattern emerges:

in all years with relatively significant numbers
(1966,1970,1973,1980,1984,1992,and 2005) with the possible exception
of 1996, there is a trend from warm water to cool water through the
year. Omitted from this are also good snowy owl years immediately
following a good year as there is evidence individual owls may return
a second winter. 1996 also stands out as it follows the several years
of very warm water in the early 1990s, but 1996 also shows a cool
down from mid-year on. So 1996 may not be an exception, but the
tail-end of a series of unusual years.

Someone who understands how to analyze climate data better than I do
may be able to undertake a more rigorous analysis.

(BTW Mike, did your note go out to tweeters, or did you respond only to me?)

To respond to Mike Denny's observation that this is a good microtine
year, the world is one big system and it seems that when you push
something down it pops up somewhere else. Climate modelers call these
teleconnections. It shouldn't be surprising to see it in operation.

--Jerry <tangren.family at verizon.net>
East Wenatchee, WA

>As much as I respect Wayne on many things, his 4-year owl cycle
>cannot be supported by any large scale dataset. He bases his
>claims on data from a single Christmas Count, the Lardner, BC
>count.
>
>Taken regionally, BC-WA-OR, there are not clear, predictable
>cycles, mega-irruptions like the one happening this year occur
>every nine to twelve years (1947, 1955, 1966, 1973, 1984, 1996,
>2005, by CBC data).
>
>I've attached 50+ years of CBC data for Snowy Owls. If you can
>find a 4-year cycle (or a statistical method for supporting one)
>pleas let me know.
>
>year "number" "reportingCounts"
>1950 0 0

...

>2003 1 1
>2004 11 5
>
>--
>Mike Patterson
>Astoria, OR
>celata at pacifier.com