Why is the projection upside down

[SimonRoss_User]

Does anyone know why the south African national grid projections display upside down in ArcMap?  ESRI appear to have deviated from the standard EPSG definitions by inverting the scale factor.  If I customise the scale factor then it displays the correct way up again.  In desktop this is fine but on the web it is causing me issues

For example I have some data that is correct if using EPSG definition 2054.  The ESRI definition with the same coordinate system but inverted scale factor is 102488.  Unfortunately my coordinate data is in a third party system and I'd prefer not to edit it.  The WKID is stored in that system as 2054.  Unfortunately the out of the box geometry server does not include 2054 as a definition so on projecting in a webmap it fails to recognise the WKID and puts the data at the same coordinate location in web Mercator (middle of the atlantic ocean)

I presume there is some specific reason why the projections have been configured this way but at the moment without either customising my geometry server or changing my coordinate data it simply doesn't work online.  I'm resigned to the fact that I probably need to make one of these changes but would really like to understand the reasoning behind this deviation from EPSG

Thanks

(Willing to bet it will be Melita that picks this one up!)

[JoeBorgione]

For sure Melita will now...  MKennedy-esristaff‌

[MelitaKennedy]

Hi Simon,

As you know, the Lo grids use positive south and west axes. We don't support those yet, everything is east-north (in that order as well).

A workaround is to change the scale factor parameter in the definition to -1. That will flip the axes so you can use the data with the base maps, etc.

Melita

[NeilAyres]

Melita, Simon,

the only real use for those, so called, "Lo" projection definitions in the Arc software is to be able to define certain "upside down" CAD files that you come across quite frequently here in SA. If you preview one of these in Catalog, generally you can see a north arrow pointing down and the extents will show a +ve Y value.

Then all you need is the zone number (17, 19, 21 etc).

But I wish esri had not named these things "Lo", cos its not. Should have called it "pseudo Lo" or something.

In reality, "Lo" is southward facing and axes swopped (X increases from N to S & Y increases from E to W) and is not supported in anything software I know about. How could it be?

Most people just convert single point coords back to a normal, right way up TM (X is Y and Y is X and both are multiplied by -1).

[NeilAyres]

And Simon,

are you sure that the geometries defined as 2054 (Lo31) are really like that. Or just normal TM that is mislabeled / defined.

[NeilAyres]

And,

having though about this for a few minutes...

If what you are seeing, when you define this data with one of those "pseudo Lo" projection in the esri software, and it appears upside down, then what you are looking at is normal TM coordinates, not real "Lo" at all.

This is the usual mis-labeling of data. Happens a lot.

See EE publishers archive from PositionIT Dec 2011, Jan & Feb 2012 for all the details.

EE PositionIT