Urban & Town Planning (CE-504 (A)) - Unit 3 Short Notes

UNIT 3: Applications & Integration of Remote Sensing & GIS

MODULE 1: FUNDAMENTALS OF REMOTE SENSING (RS)

Spectral Properties & Signatures

• Spectral Reflectance ($\rho$): Ratio of reflected radiation to incident radiation at a specific wavelength $\lambda$.

  $$\displaystyle \rho(\lambda) = \frac{E_{\text{reflected}}(\lambda)}{E_{\text{incident}}(\lambda)} $$

• Spectral Signature/Profile: Unique reflectance curve for different surface covers, used for identification.

  • Soil: Low reflectance, increases with wavelength (influenced by moisture, texture, organic matter).

    DiagramSEARCH: soil spectral reflectance curve

  • Vegetation: Low in blue/red (chlorophyll absorption), high in NIR (cell structure).

    DiagramSEARCH: vegetation spectral reflectance curve

  • Water: Low and decreases with wavelength (absorption increases).

    DiagramSEARCH: water spectral reflectance curve

• [!TIP] Spectral signatures are the basis for digital classification; similar signatures indicate similar land cover.

Remote Sensing Systems & Platforms

• Ideal Remote Sensing System: Components: Energy Source → Atmosphere-Surface Interaction → Sensor → Data Processing → Interpretation → User.

  DiagramSEARCH: ideal remote sensing system diagram

• Satellite Orbits:

  • Geostationary: ~36,000 km altitude, fixed over equator, continuous coverage of same area. Used for weather monitoring. Low synoptivity (wide-area coverage), high repetitivity (frequent revisit).

  • Sun-synchronous (Polar): ~700–800 km altitude, passes over poles, consistent lighting, wide swath (synoptivity), regular revisit (repetitivity). Used for earth observation.

• Indian Satellite Program:

  • IRS Series: Earth observation for resources, environment, agriculture.

  • Chandrayaan-3 (lunar mission): Objectives – soft landing on Moon’s south pole, rover deployment, in-situ scientific experiments. Salient features – indigenous landing system, rover Pragyan, payloads for lunar surface analysis.

• Earth Resource Satellites vs. Weather Satellites:

  • Earth Resource: High spatial resolution, multispectral, for land/vegetation/water mapping (e.g., Landsat, IRS).

  • Weather: Geostationary, high temporal resolution, for atmospheric monitoring (e.g., INSAT, GOES).

Sensor & Image Characteristics

• Resolution Types:

  • Spatial: Ground area represented by a pixel (GSD). Higher = finer detail.

  • Spectral: Number and width of wavelength bands (e.g., multispectral: 4–10 bands; hyperspectral: >100 bands).

  • Temporal: Time interval between successive acquisitions of same area.

  • Radiometric: Number of brightness levels (e.g., 8-bit = 256 levels; 11-bit = 2048 levels).

• [!TIP] Spatial resolution is often confused with map scale; it is the smallest separable ground object.

Image Processing & Analysis

• Visual Interpretation vs. Digital Image Analysis:

  | Aspect | Visual Interpretation | Digital Image Analysis | |---------------------|-----------------------------------------------|---------------------------------------------| | Method | Manual, using photo-interpretation elements | Computer-based, statistical algorithms | | Speed | Slow, time-consuming | Fast, automated | | Objectivity | Subjective, depends on interpreter | Objective, reproducible | | Output | Thematic maps, vector features | Classified raster, quantitative results |

• Elements of Visual Interpretation: Tone, texture, shape, size, pattern, association, shadow, color (for CIR).

• Digital Classification:

  • Supervised Classification: User selects training samples (ROIs) for each class; classifier (e.g., Maximum Likelihood) assigns pixels based on spectral similarity.

  • Unsupervised Classification: Algorithm (e.g., ISODATA, K-means) clusters pixels into spectrally similar groups; clusters are then interpreted.

• Image Enhancement Techniques:

  • Filtering:

    • Smoothing (Low-pass): Reduces noise, blurs edges (e.g., mean filter).

    • Edge Enhancement (High-pass): Sharpens edges, highlights boundaries (e.g., Laplacian, Sobel).

  • Contrast stretching, histogram equalization.

• Stereoscope: Device for viewing stereo pairs (overlapping aerial/satellite images from different angles) to perceive 3D terrain.

  DiagramSEARCH: stereoscope device for aerial photography

MODULE 2: FUNDAMENTALS OF GEOGRAPHIC INFORMATION SYSTEMS (GIS)

Definition & Components

• GIS Definition: \boxed{\text{A system for capturing, storing, analyzing, managing, and presenting spatial or geographic data.}}

• Key Components:

  1. Hardware: Computers, storage devices, GPS, plotters/printers.

  2. Software: ArcGIS, QGIS, GRASS GIS, etc.

  3. Data: Spatial (vector/raster) and attribute (tabular) data.

  4. People: Users, managers, technicians, analysts.

  5. Procedures: Workflows, standards, analysis methods, metadata.

• Spatial and Attribute Data Integration: Spatial features (geometry) are linked to attribute tables via a unique identifier (e.g., FID). This allows joint querying of spatial and non-spatial information.

Spatial Data Models & Structures

• TIN (Triangulated Irregular Network): Vector-based elevation model using irregular triangles; efficient for variable terrain, stores breaklines.

Data Acquisition & Input

• Sources of Data:

  • Primary: Field survey, GPS, digitizing existing maps.

  • Secondary: Published maps, remote sensing imagery, government databases (census, soil surveys).

• Objectives of Data Collection: Accuracy, completeness, relevance, timeliness, cost-effectiveness, interoperability.

• Procedure for Inputting a Map and Creating Shapefiles:

  1. Georeferencing: Assign real-world coordinates to scanned map using control points.

  2. Digitization: On-screen or tablet digitization to create features (points, lines, polygons).

  3. Attribute Assignment: Link attribute table to features (e.g., road name, type, length).

  4. Topology Building: Define relationships (connectivity, adjacency) for network or polygon analysis.

  5. Export: Save as shapefile or geodatabase feature class.

• Data Conversion:

  • Vector to Raster (Rasterization): Convert features to grid cells (e.g., polygon to cell values based on attribute).

  • Raster to Vector (Vectorization): Extract features from grid (e.g., contour lines from DEM, using edge detection).

Spatial Reference Systems

• Map Projections: Mathematical transformation of 3D Earth to 2D plane; necessary for flat maps. All projections distort some property (area, shape, distance, direction).

• Types of Coordinate Systems:

  • Geographic: Latitude/longitude (angular units), based on ellipsoid (e.g., WGS84).

  • Projected: Linear units (meters, feet), e.g., UTM, State Plane, Albers.

• UTM Projection:

  • Universal Transverse Mercator: Global system, 60 zones (6° wide). Conformal (preserves shape). Uses false easting (500,000 m) and false northing (0 m for Northern Hemisphere, 10,000,000 m for Southern) to avoid negatives. Scale factor at central meridian = 0.9996.

    DiagramSEARCH: UTM projection zones diagram

  • [!TIP] UTM is widely used for large-scale mapping; each zone has its own central meridian.

Data Quality & Errors

• Systematic Errors: Predictable, consistent bias (e.g., instrument miscalibration, datum shift, incorrect projection). Can be corrected if identified.

• Non-systematic Errors: Random, unpredictable (e.g., human digitizing error, missing data, temporary GPS signal loss). Harder to correct.

• [!TIP] Always check metadata for accuracy, source, and processing history before using spatial data.

MODULE 3: GIS SPATIAL ANALYSIS OPERATIONS

Overlay Analysis

• Concept: Combine multiple spatial layers to identify relationships and create new features.

• Types:

  • Union ($A \cup B$): Output includes all areas from all input layers.

  • Intersection ($A \cap B$): Output includes only overlapping areas.

  • Identity: Output retains all features of one layer with attributes from intersecting layer.

  • Clip: Extracts features of one layer within boundary of another.

• Applications: Land suitability (overlay soil, slope, land cover), habitat mapping, infrastructure planning.

  DiagramSEARCH: overlay analysis union intersection GIS

Buffer Analysis

• Concept: Create zone(s) at specified distance around point, line, or polygon features.

• Procedure: Select feature → set buffer distance (constant or variable) → generate polygon(s).

• Applications: Riparian buffers, noise zones around roads, service areas for facilities, environmental impact zones.

Other Analytical Operations

• Query: Attribute-based (SQL) or spatial (location) selection.

• Measurement: Calculate distance, area, perimeter, volume.

• Proximity: Nearest neighbor analysis, distance to features (e.g., schools, hospitals).

MODULE 4: INTEGRATION OF RS & GIS & URBAN/ENVIRONMENTAL APPLICATIONS

Integration Workflow

1. Preprocess RS Data: Georeference, atmospheric correction, image enhancement, co-registration for multi-temporal data.

2. Classify RS Image: Supervised/unsupervised classification to produce thematic layers (e.g., LULC).

3. Import into GIS: Georeferenced raster or vectorized classification results.

4. Overlay with GIS Layers: Combine with vector data (roads, boundaries, cadastre, DEM).

5. Spatial Analysis: Query, overlay, buffer, network analysis, modeling.

6. Output: Maps, reports, statistics, decision support.

• [!TIP] Key challenge: ensuring spatial registration (same coordinate system, resolution, and datum) between RS and GIS layers.

Land Use/Land Cover (LULC) Change Assessment Step-by-Step Procedure:

1. Acquire Multi-temporal RS Imagery (e.g., satellite images from different years, same season if possible).

2. Preprocess: Atmospheric correction, geometric correction, co-registration to align images.

3. Classification: Perform supervised/unsupervised classification for each date to produce LULC maps.

4. Post-classification Processing: Filter noise, edit classes, accuracy assessment (confusion matrix, Kappa coefficient).

5. Change Detection: Compare classified maps.

   • Methods: Post-classification comparison (most common), image differencing, spectral indices (e.g., NDVI difference).

6. Generate Change Matrix: Cross-tabulation of classes from Time 1 vs. Time 2.

   \boxed{\text{Change Matrix: } \begin{bmatrix} \text{Class}{1,t1} & \text{Class}{2,t1} \ \text{Class}{1,t2} & \text{Class}{2,t2} \end{bmatrix}}

7. Map Changes: Highlight areas of conversion (e.g., forest → urban, agriculture → water).

8. Analyze Drivers: Correlate with socio-economic data (from GIS attribute tables) to understand causes.

Sectoral Applications

• Water Resources:

  • Watershed Delineation: Using DEM from RS (e.g., ASTER, SRTM) to define drainage basins, flow directions.

  • Surface Water Mapping: Water indices (NDWI, MNDWI) from multispectral imagery to extract water bodies.

  • Groundwater Potential: Lineaments, drainage density, lithology from RS + GIS overlay with slope, soil.

  • Irrigation Management: Crop evapotranspiration (ET) from RS (e.g., SEBAL), soil moisture mapping, crop health monitoring.

• Urban Planning:

  • Urban Sprawl Mapping: LULC change detection over time to quantify expansion.

  • Infrastructure Planning: Road network planning, utility corridor mapping, green space analysis.

  • Site Suitability Analysis: Overlay of constraints (slope, flood zone, land use, geology) to find optimal locations for development.

• Traffic & Transportation Management:

  • Road network extraction from high-resolution RS imagery.

  • Traffic flow modeling using GIS integrated with real-time sensor data.

  • Congestion analysis, route optimization, public transport planning.

Problems & Challenges

• Resolution Mismatch: RS imagery (e.g., 30 m Landsat) vs. high-precision vector data (e.g., 1:10,000 map). Requires resampling, may lose detail.

• Registration Errors: Misalignment between RS and GIS layers due to inaccurate georeferencing or datum differences.

• Data Compatibility: Different formats (e.g., GeoTIFF vs. shapefile), projections, scales. Requires conversion and reprojection.

• Temporal Inconsistency: RS images from different dates/conditions (e.g., season, cloud cover) may not be directly comparable.

• Thematic Accuracy: Classification errors in RS data propagate to GIS analysis; always perform accuracy assessment.