Examples of Edge Matching Problems and Solutions

1
Examples of Edge Matching Problems and Solutions
2
First, a brief definition of edge matching

• Edge matching is defined by the University of
  Minnesotas online GIS glossary as follows
• The comparison and graphic adjustment of features
  that cross adjoining map sheets to ensure that
  the features intersect the boundary at a common,
  coincident location. A "seamless" database is
  thereby created.
• (http//www.extension.umn.edu/distribution/natura
  lresources/components/DD6097ag.html)
• It specifies the goal of seamless data and
  defines very generally the processes of comparing
  and adjusting the data to reach that goal.

3
Edge matching road datafor addressing and 9-1-1

• Well be discussing edge matching only as it
  relates to road centerline GIS data used for
  addressing and 9-1-1 call location. There are 3
  elements of road data that must be edge matched
• The lines representing the roads themselves,
• The names of the roads on each side of the
  boundary and
• The address numbers/ranges on each side of the
  boundary.
• Of course, road names and address numbers often
  change at political boundaries and can even be
  different on each side of the same road.

4
Edge matching road datafor addressing and 9-1-1

• Most of the edge matching process involves
  comparison of each road that crosses or follows a
  boundary to determine whether adjustments will be
  needed and, if so, specifically what adjustments
  are needed.
• It generally makes the most sense for local
  agencies to meet with their neighbors and make
  corrections themselves after careful examination,
  collaboration and mutual agreement. This
  requires several things that may be in short
  supply depending on those involved
• Cooperation,
• Time,
• Understanding of GIS and addressing,
• Authority to make edits to addressing and telco
  databases as needed and
• Planning for ongoing edge matching as needed.

5
Edge matching road datafor addressing and 9-1-1

• With the advent of NG9-1-1, there will be a need
  for edge matched data across the state far sooner
  than appears to be possible by organizing
  separate meetings between most GIS/addressing
  agencies around the state.
• Centralizing this work has several benefits and
  may, in fact, result in more local interest in
  undertaking the work.

6
Edge Matching Problems and Solutions

• The following slides will illustrate edge
  matching problems and combinations of problems
  that are currently known to exist in the NM 9-1-1
  GIS Database.
• For each problem, a solution has also been
  identified and illustrated. It is worth noting
  that there are generally many ways to eliminate
  an edge matching problem. The more time and
  information at the disposal of the GIS Analyst,
  the better job they will be able to do in
  reaching appropriate solutions.

7
Edge Matching Problems and SolutionsSNAPPING

• PROBLEM
• A snapping problem occurs when two roads that
  connect to each other on the ground do not do so
  in the GIS data.
• Two visible instances are shown with red circles
  to the left but please note that these types of
  problems can be too small to be seen and can
  sometimes only be identified by GIS tools.
• The danger of having roads unsnapped is that
  software will not recognize them as being
  connected.

8
Edge Matching Problems and SolutionsSNAPPING

• SOLUTION
• This is perhaps the easiest sort of problem to
  correct in that the GIS Analyst does not
  generally need to spend as long determining the
  best solution.
• They can simply correct the problem using
  standard GIS tools so that the exact location of
  the start of one segment is identical to other
  segments it is supposed to connect to. In this
  case, the corrections were made so that the roads
  connect to their neighboring features right at
  the county line.

9
Edge Matching Problems and SolutionsDUPLICATE
ADDRESSING AT BORDERS

• PROBLEM
• Although the GIS Analyst has made the necessary
  corrections to the snapping problems, there is
  also a problem with the address range attribution
  for Hobbs Hwy at the county line.
• On the next slide, both are highlighted in blue.
  The Lea version of the highway has the value
  LE_CO to the left side of the table. The Eddy
  version is marked with ED_CO.
• (continued)

10
Edge Matching Problems and SolutionsDUPLICATE
ADDRESSING AT BORDERS

• (PROBLEM continued)
• Notice how the Eddy County version ends at 6242
  in possible addressing while the Lea version
  starts at 6200. Since both have the exact same
  name and are in the Carlsbad community, that
  means that any address from 6200 to 6242 could
  map to either side of the county line.

11
Edge Matching Problems and SolutionsDUPLICATE
ADDRESSING AT BORDERS

• SOLUTION
• The GIS Analyst examined the length of the roads
  and total number of addresses available (6169
  through 6281, roughly 112 addresses). They
  determined that the Eddy feature accounted for
  72 of the length of the two combined and,
  therefore, assigned 72 of the available
  addresses to that feature (80 addresses). They
  edited the high values of the Eddy feature
  accordingly then edited the low values of the
  Lea feature so that it picks up where the Eddy
  features address range left off. The red
  circles show the changed values.

12
Edge Matching Problems and SolutionsFEATURE
OVERLAP

• PROBLEM
• Data Sources often submit road features that
  completely or partially overlap features in the
  neighboring data.
• In this case, De Baca County has submitted
  portions of Urton Lake Rd and Salt Lake Rd within
  Chaves County.
• (Note also that the 2 Data Sources disagree about
  which road is Urton Lake Rd.)

13
Edge Matching Problems and SolutionsFEATURE
OVERLAP

• SOLUTION
• The GIS Analyst removed the overlapping De Baca
  County features (as they were in Chaves County)
  and then snapped the features together at the
  county line (see the red circles).
• They also investigated the road name disparity
  using publicly-available data (GoogleEarth in
  this case) and found that the Chaves County data
  had the names switched for Sea Rd and Urton Lake
  Rd. After first checking the address ranges
  appropriately, they switched the names.

14
Edge Matching Problems and SolutionsFEATURE
SPLITTING and ADDRESS RANGING ON BORDER ROADS

• PROBLEM
• This is a classic example of the sorts of
  problems found along county line roads. In this
  case, Otero County and Dona Ana County each have
  slightly different versions of the road and each
  is responsible for its own side of that road.
• It is a standard GIS rule to always start a new
  feature at each intersection with another feature
  and Dona Ana County has done that along County
  Line Rd, splitting it at Rocky Mountain Rd. (See
  the red circle.)
• (continued)

15
Edge Matching Problems and SolutionsFEATURE
SPLITTING and ADDRESS RANGING ON BORDER ROADS

• (PROBLEM continued)
• The table below shows original attribution of the
  2 Dona Ana features. Notice how they have
  included address range data for both sides of the
  road despite only being responsible for the
  west/left side. It would have been more
  appropriate here to put zero values in for the
  right-side address range the typical convention
  for this situation.

(continued)
16
Edge Matching Problems and SolutionsFEATURE
SPLITTING and ADDRESS RANGING ON BORDER ROADS

• (PROBLEM continued)
• Otero County has not taken into account the
  existence of Rocky Mountain Rd on the other side
  of the county line and has therefore not split
  their version of County Line Rd at that
  intersection. (One road feature extends from
  Lisa Dr to Old Laredo Rd.) They have zeroed
  out the left side address range, however.

(continued)
17
Edge Matching Problems and SolutionsFEATURE
SPLITTING and ADDRESS RANGING ON BORDER ROADS

• (PROBLEM continued)
• Of course, this picture also shows a disparity in
  the location of the road but it is relatively
  small see the gap between the two versions of
  County Line Rd (shown in white).
• It is also interesting that the two sources dont
  agree on the extent of this road as Dona Ana has
  it going further north than Otero and Otero has
  it going further south.

18
Edge Matching Problems and SolutionsFEATURE
SPLITTING and ADDRESS RANGING ON BORDER ROADS

• SOLUTION
• Before splitting the Otero feature at the
  intersection with Rocky Mountain Rd, the GIS
  Analyst first made spatial edits to make it match
  the Dona Ana version of the road perfectly (and
  snap to both datasets). The Dona Ana data
  lined up better with aerial photos of the area.
• The Analyst did not attempt to move both features
  to follow the county line as shown because they
  were aware that the road data more accurately
  depicts the county line.
• It is appropriate in this case for each Data
  Source to have a feature in their data that
  represents one side of the road as they are each
  responsible for their own side.
• (continued)

19
Edge Matching Problems and SolutionsFEATURE
SPLITTING and ADDRESS RANGING ON BORDER ROADS

• (SOLUTION continued)
• The GIS Analyst split the edited Otero version of
  the road into two at Rocky Mountain Rd and the
  result is that the Otero and Dona Ana versions of
  those 2 blocks (shown in light blue) are stacked
  on each other so neatly that you cannot see that
  there are 2 features there. They also split the
  original Otero address range (600-650) in half as
  shown below