Wednesday, April 21, 2010

A little migration

(And some cynicism about migration prediction hype).

I did a little birding this morning in Princeton before work. Migrants were scarce and limited to a nice singing Blue-gray Gnatcatcher and a female Eastern Towhee. The brief lack of familiarity with the Gnatcatcher song makes me think I should dig out the spring bird song DVDs (or the playlist on the iPhone) before I get surprised again.
The pond has retreated to a puddle and but there are some optimistic Tree Swallows hanging around.

Now for your regularly scheduled cynicism. A post on JerseyBirds this morning cited

"...the floodgates indeed were opened last night-
and New Jersey experienced the heaviest migration event this season..."

from the person that runs Nice hype, not borne out with a reality in Princeton or on Staten Island based on early reports and there appeared to be a lack of epic movement in Central Park although there was certainly some. But my pet peeve about this sort of rhetoric is that given the poster is a scientist with a PhD (as I am) a lot of these reports are associated with hype and no attempt at quantitation. As such they barely survive the BS filter, although I've got expectations that there's a decent correlation between radar reflectivity and migration. Quantitation is necessary if this sort of migration detection/prediction is to get much above the level of voodoo, because as it stands I usually do at least a good job at qualitative migration prediction from checking the overnight wind direction or forecast high temps.

As for actually making predictions, the weatherunderground and Accuweather diverge on wind direction forecasts - if you believe the former there's a predominant west-to-southwest flow over the area in the next few days which augers well for migration. Accuweather's wind direction prediction is much more variable and murky. So let's invoke the tie-breaker via the
NOAA forecast which suggests potential southerly winds overnight but swinging around to north-west for at least two days after that.

No comments: