Graeme Cant wrote:

Robert Ehrlich wrote:
Mike Lindsay wrote:
...
Presumably by integrating all the information it's getting from muscle
tension in wing muscles, airspeed from pressure on feathers, and sense
organs in the lining of bone cavities?


No, integration can't work in the long term because of the
accumulation of errors along time.

No Robert. I didn't mean 'integration' and I don't think Mike meant
'integration'. We meant integration.

Our integration is not the mathematical opposite of differentiation.
We're simply saying it takes data from all its information sources to
decide what's happening. It doesn't look at each one separately like we
shift from ASI to horizon to vario to altimeter to compass picking up
separate bits of information. With a bird it's all one. It's integrated.

This is how inertial navigation
systems work, but the integration process is carried with a precision
that no brain can meet, and they need absolute attitude information
from gyros, and nevertheless need recalibration after some time. I
think without the gyros, i.e. obtaining attitude also by integration
of rotational accelerations, the precision would be lost after a few
minutes.

Bird fight is not like our flight where what we're doing is totally
alien and has to be (mostly) a conscious process. Birds fly quite
UNconsciously. What they see is "integrated" with what they feel in
their muscles. There is no separate process for "I am entering a
thermal" which is different from "I am in a steady climb". It's all one.

They measure altitude with their eyes as Bill Daniels said. They
'recalibrate' their gyros continuously by looking around. They're not
much better at raw data IMC than we are and lose their 'calibration'
quite quickly in cloud.

Many of the ideas are interesting but I'm with Occam's Razor. Birds fly
like we walk - by looking around to 'calibrate' the inner ear balance
mechanism and feedback from muscles. I can't see a real need for any
other unique mechanism.

... until we get to intercontinental migration.

GC

OK, you may call as you like the process of determining climb rate from
acceleration or muscle tension or any other differential information,
it will nevertheless be a mathematical integration, or an equivalent
process, but as the result is unique starting from a known state, any
equivalent process is the same process.

None of the mechanism mentionned in the previous post (acceleration,
muscle tension, ground watching), except the use of a pressure sensor,
can explain how birds feel they are climbing, and as a matter of fact
someone said in a previous post that the evidence of such a sensor
in the birds has been proved and also that loosing this specific
information makes them unable to climb.