Vector Data Analysis: Topology, Overlay, Network & Buffer Techniques

Core Concepts of Vector Data Analysis

Vector data represents geographic features as discrete objects (points, lines, polygons) with attributes. Unlike raster data, vector analysis focuses on topological relationships (how features connect, overlap, or relate spatially) and geometric operations (e.g., buffering, overlay). This unit covers four key techniques:

1. Topology in GIS

Topology defines how vector features spatially relate to each other (e.g., adjacency, connectivity, containment). It ensures data integrity during editing and analysis.

• Key topological rules:
  • Adjacency: Two polygons share a common boundary (e.g., Kathmandu and Lalitpur districts).
  • Connectivity: Lines (roads, rivers) meet at nodes (intersections, junctions).
  • Containment: A polygon (e.g., a village) lies entirely within another (e.g., a district).

graph TD
    A["Polygon A (District)"] -->|"contains"| B["Polygon B (Village)"]
    C["Line 1 (Road)"] -->|"connects to"| D["Node (Intersection)"]
    D --> E["Line 2 (Road)"]
    B -->|"adjacent to"| F["Polygon C (Village)"]

Why topology matters:

• Prevents gaps or overlaps in datasets (e.g., a road network where two streets meet at a node).
• Enables advanced analyses like network routing (e.g., Pathao’s delivery optimization) or territorial disputes (e.g., land boundary conflicts).

Shows adjacency, connectivity, and containment rules in a vector dataset. (Image: Arbeck, CC BY 4.0, via Wikimedia Commons)

2. Spatial Overlay Analysis

Overlay combines multiple vector layers to answer "what’s here?" questions. Common methods:

• Intersection: Identifies features common to all input layers (e.g., "Which areas are both flood-prone and near schools?").
• Union: Merges all features from input layers (e.g., combining district boundaries with election zones).
• Identity: Retains features from one layer that overlap with another (e.g., "Which roads pass through protected forests?").

Worked Example: Kathmandu Traffic Congestion

Problem: Identify high-traffic roads near schools to plan bypass routes. Steps:

1. Input Layers:
   • Layer 1: Road network (vector lines) with traffic volume data.
   • Layer 2: School locations (points) within 500m of roads.
2. Overlay: Use a buffer (500m radius around schools) + intersection with roads.
3. Output: Roads overlapping the buffer are flagged for congestion analysis.

flowchart LR
    A["Road Network Layer"] -->|"Buffer 500m"| B["School Buffer Zones"]
    B -->|"Intersection"| C["High-Traffic Roads Near Schools"]
    C --> D["Propose Bypass Routes"]

3. Network Analysis

Analyzes connected vector data (e.g., roads, rivers, utilities) to solve pathfinding, connectivity, or service-area problems. Key operations:

• Shortest path: Finds the quickest route (e.g., Pathao’s delivery optimization).
• Service area: Defines regions accessible within a time/distance (e.g., "Which areas are within 30 minutes of an NTC substation?").
• Closest facility: Locates nearest services (e.g., "Find the nearest blood bank during an emergency").

Real-World Example: Daraz Delivery Routes

Daraz uses network analysis to:

1. Model roads as a graph (nodes = intersections, edges = road segments with travel times).
2. Apply shortest-path algorithms (e.g., Dijkstra’s) to optimize delivery routes.
3. Avoid traffic hotspots (e.g., roads near schools during school hours).

4. Buffer Analysis

Creates zones at a fixed distance from features (points, lines, polygons). Used for:

• Proximity analysis: "Which villages are within 2km of a river?" (flood risk).
• Regulation zones: "No construction within 100m of a protected forest."
• Service areas: "Which areas are within 5km of an eSewa kiosk?"

Worked Example: NEPSE Stock Exchange Zones

Problem: Identify high-value properties near the NEPSE building for potential acquisitions. Steps:

1. Create a 1km buffer around the NEPSE building (polygon).
2. Overlay with a land-value raster layer (from NEB).
3. Extract properties with values > Rs. 5 million within the buffer.

Comparing Vector Analysis Techniques

In the Real World

1. eSewa & Khalti:

   • Use buffer analysis to locate merchant kiosks within 1km of high-density user areas (identified via mobile data).
   • Network analysis optimizes payment routing to reduce transaction delays.

2. NTC (Nepal Telecommunications):

   • Overlay analysis combines terrain data (raster) with tower locations (vector) to predict signal coverage gaps.
   • Topology ensures no overlaps in tower service areas to avoid interference.

3. Daraz Logistics:

   • Shortest-path algorithms (network analysis) reduce delivery times by 20% by avoiding congested routes near schools during peak hours.
   • Buffer zones around warehouses define safe loading/unloading areas.

4. Nepal Police Traffic Management:

   • Overlay of accident hotspots (points) with road networks (lines) identifies high-risk intersections for traffic lights.

Advantages and Limitations

Advantages:

• Precision: Vector data is exact (e.g., a road’s exact length vs. raster’s pixel approximation).
• Scalability: Works from local (e.g., a village map) to national (e.g., Nepal’s topography).
• Attribute-rich: Each feature carries metadata (e.g., a road’s width, material, or traffic rules).

Limitations:

• Complexity: Topological errors (e.g., dangling nodes) require manual fixes.
• Data volume: High-resolution vector datasets (e.g., LiDAR-derived contours) consume storage.
• Overlay cost: Combining large layers (e.g., Nepal’s 753 municipalities) is computationally intensive.

Exam Tip

1. Diagrams are mandatory: Always draw a Mermaid flowchart or labelled diagram for overlay/network/buffer questions. Examiners deduct marks for missing visuals.

   • Example: For a buffer question, show:
     • Input feature (point/line).
     • Buffer zone (dashed circle).
     • Overlay result (highlighted area).

2. Real-world tie-ins: Link answers to Nepalese contexts (e.g., "Use NTC’s tower placement as an example of network analysis").

3. Topology traps:

   • Avoid assuming features are topologically clean. State: "Assuming the road network has no dangling nodes..."

4. Units matter: Buffer distances should be in meters/km, not pixels (raster units).

5. Common mistakes:

   • Confusing union (all features merged) vs. intersection (only overlapping features).
   • Forgetting to project data into the same coordinate system before overlay (e.g., UTM Zone 44N for Nepal).

Key Formulae to Memorize:

1. Buffer distance = $\text{Radius}\times \text{Spatial Units}$ (e.g., 500m buffer).
2. Network cost = $\sum (\text{Edge Length}\times \text{Travel Time})$ (for shortest-path calculations).
3. Overlay area = $\text{Area of Intersection}=\int \text{Common Boundary}dx$ (simplified as "shared polygon area").