<< Part 1    Part 3 >>   Part 4 >>

Building a Transport GIS in Manipur
By Keith R Johnson 

PART 2

Method of Digitizing
The main methods of digitizing are

• Using A Digitizing Tablet (A0, A1 or A2),

• Heads-Up i.e. displaying a registered raster image on the computer screen and then using a mouse trace a road or river or to pinpoint the centroid of a village or to pinpoint the location of a bridge or to specify the an area of say a District.

• Manually from map sheets such as the 1:50,000 or 1:250,000 (usually limited to point objects)

• Output from a handheld GPS. (Global Positioning System).

The Department of Earth Sciences, University of Manipur when they have undertaken digitizing for us has mostly used an A0 digitizing tablet. Tables (layers) digitized in this way were the transport network in NE India and 100 meter contours in the Imphal Valley and surrounding foothills. Before starting the digitizing the features - roads, contours are often traced and the tracing sheets placed on the A0 digitizing tablet. In this way the information such as contours or rivers from a number of 1:50,000 toposheets can be consolidated on one A0 tracing.

Most of the geographical data in the GIS was generated using the Heads-up Method. A big advantage is that with a notebook this method can be deployed any where - office, home, airport waiting room, in the garden or near the beach. Furthermore every computer in an office can be a "digitizing work station". It is also possible to digitize initially by viewing toposheets at say a scale of 1:100,000 to get the GIS up and running and the refine the digitizing by viewing at a scale of between 1:25,000 and 1:50,000. Of course the ability to use this method requires maps to be scanned. The valley toposheets have been scanned using large format commercial scanners and the other maps available to us using a Logitech "Freescan" Scanner. Using the latter method toposheets are scanned as three strips. Large hand drawn or annotated blueprints can also be readily scanned.

The information that was initially digitized manually from toposheets were the locations of bridges and culverts - objects were assigned unique identifiers of the form G16.85 i.e. the 85th bridge/culvert identified on toposheet G16. It must be added that this process also served to familiarize staff to the co-ordinate system.

A hand-held GPS displays X and Y co-ordinates, altitude and trip distance. Whilst out on site the above information (excluding altitude which I have found to be unreliable in any event) can be automatically recorded as waypoints and downloaded back in the office. This method has been used to record the location of photographs - road condition, adjacent land coverage and the precise location of potential quarry sites. It has also been used to convert data such as roughness and strip plans that are recorded on a chainage basis to road links in the GIS. (Another application is to use the tracking facility to record the alignment of say, a road through the hills - the BCEOM team working on the Imphal - Ukhrul Road Project used this facility.)

Road Network  
The road network is the "raison d'etre" of the GIS. Specifying the road network prior to digitizing was commenced on the first day of the study - end of May 1999. The source material was the list of road sections comprising State Highways, Major and Other District Roads together with a blue print (approximately 1:250,000 scale) with roads hand-colored - National Highways (NH) - black, State Highways (SH) - red, Major District Roads (MD - blue, and Other District Roads (OD) - green. The map also contained some Inter-Village Roads (IV) - yellow. The same color coding scheme was adopted in the GIS until recently when Inter-Village Roads are now shown with a black dashed line to facilitate photocopying the A1 size Manipur and individual district maps. The colored blueprints did not show the precise specification of all the road sections on the list.

Being blueprints the base maps had become distorted over time. Furthermore the alignment of the roads was imprecise - one only has to look at the alignment of roads on the 1:250,000 Map Sheets compared to the same roads on the 1:50,000 toposheets.

The first task was to sub-divide the specified Road Sections into links. Each link was allocated a Unique Link Identifier referred to as the LinkID. e.g. SH10.12 a link that is part of the Road Section numbered 10 in the list of State Highways. For example links making up SH10 were allocated LinkIDs - SH10.10, SH10.12, SH10.14 …………… SH10.24 from start of the section (SH10.10) to the end of the section (SH10.24). The numbers start at 10 and increment by 2; incrementing by 2 allows an additional link to be inserted between two original links without affecting the sequence.

The links in the road network are stored in two tables - MAINroads - NH, SH, MD, OD and IVroads - InterVillage Roads. Nodes have been specified at the end of or at the intersections of roads in the Main Road Network. Nodes are not specified on Inter-Village Roads. The fields in the MAINroads table are: -

• DIST - District

• CLASS - Class (NH, SH, MD, OD)

• RS - Road Section

• RSdesc - Road Section Description

• RSlength - Road Section - Length (as specified on the list of roads)

• RStype - Road Section - type of construction

• RScond - Road Section - condition of road (from PWD records)

• RSwidh - Road Section - road width

• LinkID

• NA - node number at start of link

• NB - node number at end of link

• LEN - Length of link derived from the GIS. (using the function SphericalObjectLen)

• GPSlen - Length of Link derived from GPS outputs from a drive along the Road Section)

• Terrain - Type of Terrain

• Flow - Traffic Flow

• Capacity - Link Capacity

• Rwidth - Road Width (from Strip Plan Survey where available)

In the table IVroads the fields RS, RSdesc, RSlength, RStype, RScond, RSwidth are dummy variables and are blank. As nodes are not specified on Inter-Village Roads NA and NB are also dummy variables and are blank.

The initial link digitization was based on 1:250,000 scale blueprints, this was later refined using individual district blue print maps. These maps proved to be inconsistent and the current digitization is based on the 1:50,000 toposheets The roads having been drawn on photocopies of the toposheets .by the PWD Engineers with the most detailed knowledge of the districts. Checking and amending road specifications is a continuous process. The Road Network is shown in Figure 2 and the detailed coding in Figure 3.

An additional table - LinkIRI has been created from the Bump Integrator Survey, which was carried out along the Road Sections selected for the Feasibility Study on a chainage basis - one estimate of IRI for each kilometer was recorded. In LinkIRI the average IRI for the link together with range of IRI values along the link are stored.

A further table STRIPplans relates Strip Plan Drawing Numbers to LinkIDs.

Back to Top              << Part 1    Part 3 >>   Part 4 >>