In article , Mike Tullett writes:
On Tue, 31 May 2005 14:15:15 +0000 (UTC), Keith Dancey wrote in


http://www.bbc.co.uk/weather/ukweath...e.shtml#no_url

Ever since the new graphics were introduced these satellite charts have
had the imagery displaced about 150 miles north of where it should be.
It's a constant error. The satellite imagery never matches up with the
radar imagery.

Parallax error from geostationary view, projected onto image frame and
rotated forward into a map projection (ie no correction for difference
in altitude between surface and cloud, during the forward rotation)?

It's still like it tonight. They seem to be using two different
projections, one for the cloud and the other for the land. Then they
superimpose the cloud onto the land and the mismatch is seen - or is that
what you are saying:-) ?


What I *think* may be happening, to cause this constant displacement, is that
the BBC are taking satellite visible images (which will show a parallax error
between cloud and land from the geostationary POV) and then rotating the
whole image forward (pitch rotation about the z-axis of around 70 degrees)
as if the entire image lay in a single plane, in order to *attempt* to
display the image as seen from directly above.

The BBC have no scientific or mathematical credentials, and such a manipulation
of the original image will cause two gross errors: it will distort the shape
of the image because no account is taken of the curvature of the Earth and of
perspective, and it will show a constant displacement of the cloud with respect
to the surface because they have no altitude information with which to correct
the original parallax error?


Maybe that also explains why the "new-new" BBC weather maps - with an apparent
view from above - *appears* to distort Scotland. If they are simply applying a rotation to the perspective view (obtained from a geostationary POV) without
correcting for curvature of the Earth or for perspective, then the resulting
image will be distorted from reality.

(I say "*appears* to distort Scotland" because I have not yet looked at the
new projection to judge whether I think Scotland is distorted, but have read
such comments here)


It is fair to say that the UKMO would/should NEVER perform such an abomination to
original data because they employ scientists, and would report their methods.



...Even if they move the "cloud" to match the land
it still leaves a question - of the grey area, just what is real cloud and
what is an artifact of the system?

Not sure I follow this... but the BBC ought to publish its methods!



I'm still bemused by these words at the top

"These satellite charts show the image you can see from space, so are only
available during the day between 1100 and 1600" :-(


We are left speculating (again): maybe their data capture, followed by
data manipulation, limits the availability of *visible* images even beyond the
normal limits of seasonal daylight?


Cheers,

keith



---
Iraq: 6.5 thousand million pounds, 80 UK lives, and counting...
100,000+ civilian casualties, largely of coalition bombing...