the birds in the trees by the canal

jinian suggested cormorants. That seems to be them, from pictures. The one that looks like the best match is Brandt's cormorant.

(I love how that enature.com page offers search choices "all species" and "dangerous".)

flanging of car noise on curved roads

Walking home: a car coming down Palatine to me at the bottom of the stairs, the intersection with 43rd. Its noise sounds flanged upwards.

Hmm.

Okay, this seems justifiable. It's that the road there has upward curvature. On a flat road, the length difference between the direct path and the bounce from the road is quite small -- direct path is hypot( distance, height_car - height_person ), and bounce (reflect the virtual person-ear below ground) is hypot( distance, height_car + height_person ) -- so maybe we don't notice it shifting? I haven't anyway. Huh, and actually it gets longer as the car comes closer, for a downward flange.

Whereas with a curved road, say it's a circular arc, the bounce path is substantially longer than the direct -- hypot( distance, 2 * dip ) - distance, where dip is how far the road dips below the straight-line path through the air, is the approximation for dip >> height_{person,car}. And as the car comes closer it gets less so.

What's the road such that the length difference is constant? Will check back tomorrow.