### A paper and Pencil Exercise in Digitizing

Game Plan:

Using the map excerpt above (from Bryan's Sectional Maps O'AHU ) and a sheet of 6, 8 or 10 divisions to the inch tracing graph paper, plan to "digitize" the roads (centerlines) and streams two ways.

Vector Style: On the graph paper, add axes and an origin. Align the axes to the left and below the map excerpt. Select points to capture the shape and topology of the lines and mark (trace) them on the graph paper. Mark the points (vertices) along the lines with dots (.) and mark the topological intersections (nodes) with asterisks (*).

Measure the length of the road and stream lines. Number the arcs (lines from node to node) and the nodes. Count the number of points, nodes, and arcs in both networks. Report them in a table like:

```...........   Roads    :    Streams
Length (cm)
nodes
arcs
points
```

Number the vertices in sequence along about two inches of the linework, and determine grid coordinates (to one decimal place if you like) for those points. Then, list the arc #, point sequential #, and X and Y coordinates for those vertices in a table with these columns:
Arc # : Point # : X-cood : Y-coord

Raster Style: Again, with the map excerpt and graph paper, lay out a raster with rows numbered from top to bottom and columns from left to right, and position it with the origin above and to the left of the map. Build a raster model of about a 1 inch by 1 inch area of the map, coding the roads and streams using integers empty ('0'), road ('1'), road intersection ('2'), stream ('4'), stream junction ('8') --- and summing those to code cells that have multiple attributes, so, a stream crossing a road is '5' (=1+4) and a road coinciding with a stream junciton is '9' (=1+8). List the first 30 values of your model in explicit row order:
Row : Column : Value

Finally, a little reflection: In each case, what decisions did you need to make in order to do the coding? Were there situtions with more or less certainty in the accruacy of the coding? How would the resolution of the coordinate grid (coarser or finer graph paper) impact the accuracy? The work? The data volume?