Registration and georeferencing
Registration: lining up layers with others in a dataset
Geo-registration (georeferencing): linking a layer or dataset with spatial coordinates
If the world was flat, we could use a simple coordinate system. with 0.0 in the bottom left corner. But its NOT.. (DOH!! )
Ground coordinates are based on a surveyed ellipse that approximates the Earth. This ellipsoid has slightly different equatorial and polar radii, due to 'polar flattening'. Although it is only about 1/300th of the earth's radius, the calculation of this ellipsoid over the last century has given different ground coordinates. Georeferencing-Geog205
Today, there is agreement on the World Geodetic System 1984 WGS-84
In North America there have been two standards (North American Datum = NAD)
NAD27 versus NAD83: the 'NAD shift' (North America)
There can be a difference (NAD27-NAD83) of up to 170 metres in y (N-S) and 70 metres in x (E-W), corresponding to the difference between the two ellipsoids. This is known as the 'NAD shift'. GIS software can make this conversion, otherwise data based on the two different NAD standards will not 'match' or overlay at large scales.
Problems in using the graticule in GIS:
Although geographic data are sometimes referred to as 'unprojected', the format corresponds to the Plate Carree
Geographic is a suitable system for storing global data, but NOT for display or analysis
e.g. the Natural Resources Canada toporama website (PG map sheet is 93G or 93G15 (1:250,000 / 1:50,000)
A map projection is a method of showing the 3-D planet Earth on a 2-D plane, (inevitably causing distortions). The distortion can occur in distance, area, shape, or direction. Each projection will result in different types and amounts of distortion.
The 3 main types of projections are: shown in this graphic
Generally, there is least (none) distortion where the globe touches the projection surface, but distortion increases away from this point or line(s)... and to further complicate matters, these can be modified by projection orientation
For more information on map projections, go to the Geography 205 lecture on Map Projections
The most common georeferencing system in Canada, used on our topographic maps is the Universal Transverse Mercator (UTM) system, based on the Transverse Mercator projection (below).
The most common system in Canada is the Universal Transverse Mercator (UTM): the basic details are:
UTM data for multiple zones
For analysis of areas that cross 2 or more zones, data must be re-projected into just one zone (as if they were in that zone) ..topographic map example ... although eastings coordinates will go further away from the central meridian (500,000) and potentially below 0 .. and area distortion will increase.
View the coordinates on mapplace website to see UTM coordinates over multiple zones
Provinces slightly more than one UTM zone may use a modified system:
Alberta crosses 2 zones 11-12 (110-120W) - and uses '10TM' (one 10 degree zone centred on longitude 115 W)
Saskatchewan crosses 2 zones 12-13 (102-110W)
Manitoba crosses 2+ zones 13-15 (89-102W)
For this reason, BC chose the Albers equal area projection for storing and displaying provincial data.
Albers is a conic equal-area projection usually with 2 standard parallels:
BC Albers projection BC GIS website
This gives one set of coordinates with zero area distortion for any part of BC Albers map and grid
Other locations might use a similar projection (North America) but with a different centre and parallels
The Canada Albers Equal Area projection: central meridian at 96W, origin at 40N and standard parallels at 50, 70N
Yukon : Yukon Albers Projection - note the different coordinates and standard parallels
Canada: Lambert conformal conic - note different parallels again and it is conformal (shape), not equal area
So you are likely to see coordinates in any of these three systems for BC GIS data:
What approximate values would Prince George be in each of these three systems above? see lrdw - imap website
Identifying coordinate systems
Note that you can receive data without proper documentation, and might need to guess (and which is x and y)
Define projection: relabels the data (adds 'metadata') but changes ONLY the description (not the data)
Reproject: changes the data projection coordinates (e.g. from UTM to Albers, or NAD27 to NAD83)
Ideally all layers should be maintained using the same coordinate system .. although if properly defined (labelled), software can reproject layers 'on the fly'.. otherwise layers will not overlay and in any case cannot be combined for analysis.
Q: If layers are from the 3 different systems above, and not properly defined, how would they locate relative to each other?