Correctly georeference rainfall and cloud forecast layer

[Andrew Tyrrell]

I have accessed the rainfall and cloud forecast layers using FME, and have attempted to georeference the images using the info provided on the Datapoint website:
"The map layer is provided without a map, the boundary box for this image is 48.0 to 61.0 degrees north and 12.0 degrees west to 5.0 degrees east.
The image layers are currently made available in a Mercator projection"

I have set the initial coordinate system to WGS1984, and specified the corner coordinates of the image to:
X Upper Left Coordinate: -12
Y Upper Left Coordinate: 61
X Upper Right Coordinate: 5
Y Upper Right Coordinate: 61
X Lower Right Coordinate: 5
Y Lower Right Coordinate: 48
X Lower Left Coordinate: -12
Y Lower Left Coordinate: 48

I have then reprojected this into British National Grid and overlaid a shapefile of the coast, to check the position.

At first glance, it seems that the image has projected into the correct position.  However, when comparing it to the layers available in the forecast page of the Met Office website, it seems that the georeferencing is actually slightly off.  Please see the below examples:

To me it looks like a rotation issue, suggesting that the coordinate system is incorrect (the actual extents of the images seem quite accurate, so the coordinates seem correct).

Could someone please let me know what coordinate system the initial coordinates should be in?  Or alternatively, the British National Grid extents of the image?

I would very much hope to see georeferenced images available in the revamped Datapoint when it goes live, or at least some more detailed metadata about the extent/coordinate system of the data.

[Andy Turner]

Hi,

Did you resolve this issue?

Thanks in advance,

Andy