ArcINFO Overlay Methods for a Street Tree GIS

The purpose of this document is to provide some standard to street tree inventory methodology in Portland Oregon.† Following this method will assure consistency with PSU mapping efforts in the Irvington, Forest Grove, Rose, Ports, North, Cathedral, and St. Johnís neighborhoods.† These inventory projects all involved data placed into a GIS.† The sequence of GIS steps described creates a street block geography that can be used to place street trees as points that follow a grid lines each connected by nodes to form a geometric block.†

Each of the block faces can be populated with tree locations measured from a block corner and in the topology of these data, a NODE location.† Node identification numbers are a system of FID and TID identification numbers that run through the entire neighborhood from north to south.† Street corners are digitally tagged with numbers range in value from 1 at the northern top of each neighborhood, ascending to the south in increments of one.† These numbers accumulate as records in a data file accounting for each street corner in the study neighborhood.†

Within the polygon topology, each block face can be conceptualized as a line being between one corner (FID) and another corner (TID).† Identification numbers are derived according to a line data file (AAT) which is also used to identify street corner locations.† The connection between field data sheets and data files are provided by FID and TID unique identification numbers.† The collected inventory data can be placed as points between the FIS and TID locations that represent block corners.†

The GIS processing requires two basic METRO (RLIS) coverages.† The first is a centerline file.† This layer will be used to create the block geometry and to populate the neighborhood street corners with identification numbers.† The second layer is a neighborhood (polygon) boundary that defines the outer extent of the study neighborhood.† Using GIS tools these file can be changed to conform to the demands of the PSU model for street tree inventory.†

BUFFER the neighborhood polygon layer by using a 50 foot buffer distance. The new buffer layer will be used as the block geography.

CLIP the METRO centerline file with the new neighborhood buffer.

Use the Arc Info editor ARCEDIT to trip and clean the dangles out of the new clipped centerline file.

BUILD the new centerline file.

Use NODEPOINT to create a point layer of new centerlines nodes.

BUILD the new point layer.

BUFFER the new point layer with a buffer distance of 25 feet.† This distance will become important for the first overlay between buffered coverages.

BUFFER the new centerline file with a 15 foot buffer distance.

INTERSECT the two (15í and 25í) buffer coverages using the polygon option and a .01 fuzzy tolerance.† This step will create block corner locations by capturing the inner corners from the 15í centerline buffer and placing them within the 25í point buffer polygons.† The resulting coverage should contain a grid of circles with 2-4 notches spaced in quadrants around the inside perimeter of each circle.†††††

CREATELABLE for the new intersect coverage.† This step assures that all polygons contained in the coverage have labels.

CLEAN the new intersect coverage with a .01 dangle and a .01 fuzzy tolerance.† Cleaning the coverage reconstructs a relationship between the spatial components internal identification system and a link to the spatial attribute file.

BUILD the new intersect coverage as a polygon.† Building a coverage freshens the relationship between the internal identification system (with the spatial attribute file) and† the coverages attribute table.

ELIMINATE the new intersect cover by selecting area greater than 60 square feet.† The eliminate process will get rid of the larger circle connecting the intersected block corners.††

CLEAN the newest eliminate coverage with a .01 dangle and a .01 fuzzy tolerance.

BUILD the newest eliminate coverage as a polygon.

UNION the new eliminate (circle) coverage with the new line buffer coverage using a one foot tolerance.† This step combines the street corner locations with the block geography of the neighborhood.† The new union coverage contains all of the components needed to place identification number locations on the block corners.

ELIMINATE the new union coverage by selecting the AREA less than 60 square feet.† This step selects and deletes the sliver polygons that can be found at each block corner of the new union coverage.

BUILD the newest eliminate coverage as a line.† Building the new union coverage as a line eliminates the need to clean the coverage.† Excessive cleaning can move features contained within the layer.

COPYCOV the newest eliminate coverage to save a copy of it just in case itís needed to solve editing problems.

ARCEDIT the newest eliminate coverage to assure each block corner has one node.

BUILD the newest eliminate coverage as a line.

Repeat the ARCEDIT with the newest eliminate coverage to assure a high degree of quality control.

BUILD the newest eliminate coverage as a line.

The final coverage can be used as a base layer for planning the collecting data about street trees because of the spatial relationship between the block faces and the FID TID identification numbers.† Plotting the coverage with NODE ID on the block corners and street names from the new (clipped) centerline file is very useful for orienting field data collectors.†

This methodology can be carried out in just a few hours from beginning to end.† Editing node errors in ARCEDIT is the most time consuming part of the data processing.† Nodes at intersections can be missing as a result of road intersections not forming 90 degree angles with each other.† Arcs can be selected and split at intersections with missing nodes.† Other intersections could have more than one node.† In this case adjacent arcs should be selected and UNSPLIT.† In upwards of 95% of all intersections have had no node errors in street tree inventories processed at the PSU Cartographic Center.

If node errors are a problem, adjusting the FUZZY TOLERANCE in a CLEAN to a higher number.† This kind of action can move features in a coverage equal to the distance of the tolerance.† Street intersection nodes can be snapped together provided the connection arc is selected by the FUZZY TOLERANCE.† Too much cleaning can destroy the continuity of the block geography.