Remote Sensing & GIS (CE-504 (B)) - Unit 4 Short Notes

A. FUNDAMENTALS OF REMOTE SENSING (RS)

Spectral Properties & Reflectance

• Spectral Reflectance ($\rho$) is the ratio of reflected radiance to incident radiance at a specific wavelength:

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

• Spectral Reflectance Curves: Plot of $\rho$ vs. wavelength (400–2500 nm) for different surface features.

  • Vegetation: Low in visible (blue/red absorption by chlorophyll), high peak in NIR (cell structure), moderate in SWIR (water content).

  • Water: Low overall; strong absorption in NIR and SWIR; appears dark.

  • Soil: Generally increases with wavelength; varies with moisture, texture, and mineral composition.

[!TIP] Vegetation’s high NIR reflectance is key for NDVI calculation. Water absorbs strongly beyond 700 nm, enabling water body discrimination.

Remote Sensing System & Components

An ideal remote sensing system comprises:

1. Energy Source (Sun or active sensor)

2. Atmosphere Interaction (scattering, absorption)

3. Target Interaction (reflection, emission)

4. Sensor Detection (captures reflected/emitted energy)

5. Data Recording & Processing (creates usable imagery)

Sensor Resolution

Resolution defines the ability to distinguish details. Four critical types:

[!TIP] Higher spatial resolution ≠ better classification; trade-offs exist with spectral and radiometric resolution.

Satellite Platforms & Orbits

• Geostationary Satellite: Orbits at ~36,000 km, matches Earth’s rotation → fixed position over equator. High repetivity (continuous monitoring), low spatial resolution. Used for weather (e.g., INSAT, GOES).

• Sun-synchronous Satellite: Polar orbit, passes same latitude at same local solar time. High synoptivity (consistent lighting), moderate repetivity (days to weeks). Used for Earth observation (e.g., IRS, Landsat).

• Synoptivity: Ability to capture wide-area, simultaneous snapshots.

• Repetivity: Frequency of revisiting the same area.

• Indian Satellite Program:

  • IRS Series: Earth observation (Resourcesat, LISS, AWiFS).

  • INSAT Series: Meteorological and communication.

  • Chandrayaan-3: Lunar mission (2023) – objectives: soft landing, rover operations, lunar surface studies; features: indigenous lander (Vikram), rover (Pragyan), scientific payloads for elemental analysis.

[!TIP] Sun-synchronous orbits ensure consistent illumination for change detection; geostationary ideal for real-time weather tracking.

B. IMAGE ACQUISITION & PROCESSING

Image Interpretation & Analysis

Elements of Visual Interpretation:

• Tone (brightness/color)

• Texture (roughness/smoothness)

• Shape (geometric form)

• Size (relative dimension)

• Pattern (spatial arrangement)

• Association (relationship with other features)

Differentiation:

Image Classification

• Supervised Classification:

  1. Define training sites (regions of known land cover).

  2. Extract spectral signatures from training data.

  3. Classify entire image using signatures (e.g., Maximum Likelihood, SVM).

• Unsupervised Classification:

  1. Algorithm clusters pixels into spectrally similar groups (e.g., ISODATA, K-means).

  2. Analyst assigns land cover labels to clusters.

• Key Difference: Supervised uses prior knowledge; unsupervised discovers natural groupings.

[!TIP] Supervised requires accurate training data; unsupervised may produce statistically optimal but thematically ambiguous clusters.

Pre-processing & Enhancement

• Image Filtering:

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

  • High-pass (Sharpening): Enhances edges, details (e.g., Laplacian, Sobel).

  • Band-pass: Isolates specific frequency features.

• Errors in RS Data:

  • Systematic: Predictable, correctable (sensor calibration, orbital drift).

  • Non-systematic: Random, unpredictable (atmospheric haze, cloud cover).

• Corrections:

  • Radiometric: Corrects for sensor noise, atmospheric effects, illumination.

  • Geometric: Corrects for Earth curvature, platform motion, terrain displacement (using GCPs and polynomial transformation).

C. GEOGRAPHIC INFORMATION SYSTEM (GIS) FUNDAMENTALS

Definition & Components

• GIS Definition: A system for capturing, storing, analyzing, managing, and presenting spatial or geographic data.

• Key Components:

  1. Hardware (computers, GPS, scanners)

  2. Software (ArcGIS, QGIS, GRASS)

  3. Data (spatial and attribute)

  4. People (users, managers)

  5. Methods/Procedures (data collection, analysis workflows)

GIS Data: Sources & Objectives

• Sources:

  • Primary: Field surveys, GPS, digitization.

  • Secondary: Maps, satellite imagery, census data, existing databases.

• Objectives of Data Collection:

  • Support specific analysis (e.g., site suitability, network analysis).

  • Ensure accuracy, completeness, and relevance.

  • Maintain interoperability and standards.

Data Models & Structure

• Data Conversion:

  • Raster to Vector: Vectorization (tracing edges, thinning).

  • Vector to Raster: Rasterization (assigning cell values based on feature attributes).

[!TIP] Vector preferred for discrete features (roads, parcels); raster for continuous surfaces (elevation, temperature).

Spatial & Attribute Data Integration

• Integration: Each spatial feature (point/line/polygon) has a unique key linking to its attribute table (non-spatial data).

• Geodatabase: A database optimized for spatial data, storing:

  • Feature classes (spatial data)

  • Attribute tables (tabular data)

  • Relationships (links between tables)

  • Topology (spatial rules)

D. COORDINATE SYSTEMS & MAP PROJECTIONS

Coordinate Systems

Map Projections

• Why Important?: Earth is 3D; maps are 2D → projections introduce distortion in shape, area, distance, direction.

• Types:

  • Cylindrical (Mercator): Conformal, distorts area near poles.

  • Conic (Albers): Equal-area, good for mid-latitude regions.

  • Azimuthal (Stereographic): Preserves shape from a point.

• Universal Transverse Mercator (UTN):

  • Key Features:

    • 60 zones (6° wide), each with central meridian.

    • Scale factor at central meridian: \boxed{k_0 = 0.9996} (slightly reduces scale).

    • False easting/northing to avoid negative coordinates.

    • Conformal (preserves local shape).

• Distortion: No projection is perfect; choice depends on map purpose (e.g., navigation vs. area calculation).

[!TIP] UTM is widely used for large-scale mapping; always check zone and datum (e.g., WGS84 vs. NAD83).

E. GIS DATA INPUT & MANAGEMENT

Data Input Procedures

1. Digitization:

   • Manual: Heads-up (on-screen) or heads-down (tablet).

   • Automated: Scanning + vectorization (edge detection).

2. Create Shapefiles:

   • Define geometry type (point/line/polygon).

   • Set coordinate system.

   • Digitize features; attributes entered manually or joined from tables.

3. Data Editing: Modify vertices, fix topology errors.

4. Topology Building: Define spatial relationships (e.g., no gaps between polygons, connectivity of lines).

Data Quality & Errors

Problems of Using RS Data in GIS:

• Scale Mismatch: RS pixel size vs. map scale.

• Resolution Issues: Mixed pixels, coarse resolution obscuring details.

• Geometric Errors: Residual misregistration after correction.

• Thematic Accuracy: Classification errors (spectral confusion, training site bias).

• Temporal Inconsistency: Multi-temporal images may have different acquisition conditions.

[!TIP] Always assess RS data accuracy before GIS integration; use ground truth data for validation.

F. GIS SPATIAL ANALYSIS OPERATIONS

Query & Measurement

• Attribute Query: Select features based on non-spatial attributes (e.g., SELECT * FROM roads WHERE type='Highway').

• Spatial Query: Select based on location (e.g., features within 1 km of a river).

• Measurement:

  • Distance: Euclidean or geodesic (on sphere).

  • Area/Perimeter: Calculated from polygon coordinates (planar for projected data).

Overlay & Neighborhood Analysis

• Overlay Analysis: Combines multiple layers to create new features.

  • Union: All features from both layers.

  • Intersection: Only overlapping areas.

  • Identity: Features of one layer clipped by another, retaining attributes.

  • Erase: Removes areas of one layer from another.

• Buffer Analysis: Creates zone of specified distance around features.

  • Applications: Riparian buffers, noise zones, service areas.

[!TIP] Overlay in raster uses Boolean logic; in vector uses polygon overlay operations. Buffer distances must use appropriate coordinate system (projected for accurate meters).

Other Operations

• Reclassification: Assigns new values to existing classes (e.g., simplifying land cover categories).

• Surface Analysis:

  • Slope: Gradient from DEM (degrees or %).

  • Aspect: Direction of slope (0–360°).

  • Derived from DEM using neighborhood statistics (e.g., Horn’s method).

G. INTEGRATION OF RS & GIS & APPLICATIONS

RS-GIS Integration

• Role of RS: Primary source of spatial data (imagery) for GIS.

• Workflow:

  1. Acquire multi-temporal RS data.

  2. Pre-process (atmospheric, geometric correction).

  3. Classify to create thematic maps (e.g., LULC).

  4. Import classified raster into GIS; convert to vector if needed.

  5. Integrate with other GIS layers (roads, boundaries) for analysis.

  6. Perform spatial analysis (overlay, buffer, change detection).

Land Use/Land Cover (LULC) Change Assessment

Procedure:

1. Multi-temporal image acquisition: Same sensor, season, and minimal cloud cover.

2. Pre-processing: Co-registration, atmospheric correction.

3. Classification: Supervised/unsupervised for each date.

4. Change Detection:

   • Post-classification comparison: Compare classified maps; generate transition matrix.

   • Other: Image differencing, vegetation index trends.

5. GIS Integration: Spatially analyze change patterns, calculate areas, identify drivers (proximity to roads, urban sprawl).

Application Domains

Water Resources:

• Watershed Delineation: Use DEM to define drainage networks and boundaries.

• Surface Water Mapping: Identify water bodies from RS (NDWI, thresholding).

• Groundwater Potential: Integrate geology, lineaments, drainage, slope.

• Flood/Drought Assessment: Monitor inundation extent (SAR/optical), vegetation health (VCI).

• Irrigation Management: Crop water requirement mapping, scheduling.

Other Applications:

• Traffic Management: Real-time monitoring, route optimization.

• Urban Planning: Urban sprawl mapping, infrastructure planning.

• Disaster Management: Hazard zonation (landslides, earthquakes), emergency response.

[!TIP] For LULC change, post-classification comparison reduces errors from atmospheric and sensor differences between dates.

H. SPECIFIC TERMS & CONCEPTS (Short Note Topics)

Stereoscope

• Definition: Optical device for viewing overlapping aerial photographs in 3D (stereoscopic vision).

• Use: Photogrammetry for elevation extraction, feature identification, terrain analysis.

Key Point: Requires parallax from overlapping images (typically 60% forward overlap).

Spectral Reflectance Curves

• Graphical representation of reflectance vs. wavelength for Earth surface features.

• Significance: Enables spectral separation of land cover types; basis for band selection and vegetation indices (e.g., NDVI = $$\displaystyle \frac{(NIR - Red)}{(NIR + Red)} $$).

Image Filtering

• Concept: Mathematical operation on pixel neighborhoods to enhance or suppress features.

• Types:

  • Low-pass: Smoothing, noise reduction (mean, median).

  • High-pass: Edge enhancement (Laplacian, gradient).

  • Band-pass: Isolates specific spatial frequencies.

Application: Pre-processing for classification, feature extraction.

UTM Projection

• Universal Transverse Mercator: Global system of 60 transverse Mercator zones.

• Key Features:

  • Scale factor $$\displaystyle k_0 = 0.9996 $$ at central meridian.

  • False easting (500,000 m) and northing (equator 0 m for Northern Hemisphere).

  • Conformal, minimal distortion within zone.

• Use: Large-scale topographic mapping, GIS data integration.

Buffer Analysis

• Concept: Creates a zone of specified distance around a spatial feature (point, line, polygon).

• Applications: Environmental impact assessment (e.g., buffer around wetlands), service area delineation (e.g., hospitals), corridor planning.

Note: Requires projected coordinate system for accurate distance.

Overlay Analysis

• Concept: Spatial overlay of two or more layers to combine their attributes and geometries.

• Types:

  • Vector: Union, intersection, identity, erase.

  • Raster: Boolean operations (AND, OR, XOR) on cell values.

• Use: Suitability analysis, change detection, intersection of zones.

Synoptivity and Repetivity

• Synoptivity: Ability to capture a wide, contiguous area in a single acquisition (e.g., satellite swath). Enables regional studies.

• Repetivity: Frequency of revisiting the same area (temporal resolution). Enables monitoring of dynamic processes (e.g., crop growth, floods).

Trade-off: Sun-synchronous orbits offer good synoptivity and consistent lighting; geostationary offer high repetivity but coarse resolution.