Remote Sensing & GIS (CE-504 (B)) - Unit 3 Short Notes

A. FUNDAMENTALS OF REMOTE SINGING

Electromagnetic Spectrum & Radiation Principles

• Energy Source: Primarily the Sun (passive sensors) or artificial sources (active sensors like RADAR, LiDAR).

• Atmospheric Interaction: Key processes affecting radiation reaching the sensor:

  • Scattering: Redirects radiation (Rayleigh by small molecules, Mie by aerosols, non-selective by large particles).

  • Absorption: By atmospheric gases (O₃, H₂O, CO₂). Creates Atmospheric Windows—wavelength bands with minimal absorption where RS operates (e.g., visible, NIR, microwave).

• Spectral Reflectance Curve: Graph of % Reflectance vs. Wavelength for a target. Unique signature based on material's chemical/physical properties.

[!TIP] Exam Focus: Be ready to sketch and explain the characteristic spectral reflectance curves for vegetation, soil, and water. Highlight key absorption/reflection bands (e.g., chlorophyll absorption in blue/red, high NIR reflectance for healthy vegetation).

Spectral Behavior of Earth Features

1. Vegetation:

   • Low reflectance in blue & red (chlorophyll absorption).

   • High reflectance in NIR (cell structure).

   • High absorption in SWIR (water in leaves).

   • Phenology (seasonal changes) alters the curve.

2. Soil:

   • General increasing reflectance with wavelength.

   • Moisture: Decreases overall reflectance, especially in SWIR.

   • Texture: Coarser soils reflect more than fine soils.

   • Organic Matter: Darkens soil, lowers reflectance.

3. Water:

   • Low reflectance in visible & NIR (absorption increases with wavelength).

   • High in turbid/sediment-laden water (especially in red/NIR).

   • Very low in clear, deep water (appears black in NIR).

Remote Sensing Systems & Platforms

• Ideal RS System Components: Energy Source → Atmosphere → Target → Sensor → Platform → Processing → User.

• Orbit Types:

  | Feature | Geostationary (GEO) | Sun-Synchronous (SSO/Polar) | | :--- | :--- | :--- | | Altitude | ~36,000 km | ~700-800 km | | Orbit | Equatorial, matches Earth's rotation | Polar, passes over poles | | Revolution | 24 hrs (stationary over one point) | ~90-100 mins | | Coverage | ~1/3 Earth (continuous) | Global (swath coverage) | | Resolution | Low spatial (1-4 km) | High spatial (10m-1km) | | Applications | Weather monitoring, communications | Land resources, environment, mapping |

• Synoptivity: Ability to capture a wide-area view in a single acquisition.

• Repetivity: Ability to revisit the same area at regular intervals.

Sensors & Data Acquisition

• Sensor Resolutions:

  • Spatial: Minimum separable distance on ground (Pixel size).

  • Spectral: Number & width of wavelength bands (multispectral vs. hyperspectral).

  • Temporal: Revisit time/time interval between acquisitions.

  • Radiometric: Number of brightness levels (e.g., 8-bit = 256 levels). Dynamic Range = Max detectable signal / Min detectable signal.

• Active vs. Passive:

  • Passive: Detects reflected/emitted natural energy (e.g., optical sensors). Works only in daylight (except thermal).

  • Active: Own energy source (e.g., RADAR, LiDAR). Works day/night, all weather.

• Scanner Types:

  • Across-track (Whiskbroom): Mirror sweeps perpendicular to flight path. Single detector per band.

  • Along-track (Pushbroom): Linear array of detectors perpendicular to flight path. No moving mirror.

Image Interpretation & Analysis

• Elements of Visual Interpretation: Location, Size, Shape, Tone/Color, Texture, Pattern, Association, Height/Shadow.

• Digital vs. Visual Analysis:

  • Digital: Computer-based, quantitative, uses spectral information, reproducible.

  • Visual: Human-eye based, qualitative, uses spatial context, subjective.

• Image Filtering (Convolution):

  • Purpose: Noise removal (smoothing), edge enhancement (sharpening).

  • Low-pass (Smoothing): Averages neighborhood → reduces noise, blurs edges.

  • High-pass (Sharpening): Highlights edges, enhances details.

  • Directional (Edge Detection): Enhances features in specific orientation (e.g., Roberts, Sobel).

• Stereoscopy: Use of stereoscope to view overlapping aerial/satellite images (stereopair) from different angles to perceive 3D depth and interpret terrain/elevation.

B. SATELLITE SYSTEMS & INDIAN SPACE PROGRAM

Earth Observation Satellite Types

1. Earth Resources Satellites:

   • Landsat (NASA/USGS): Multispectral (e.g., OLI, TIRS). Applications: LULC, agriculture, forestry, geology.

   • Sentinel (ESA): Sentinel-2 (multispectral, high res), Sentinel-3 (ocean/land). Applications: similar to Landsat, with higher temporal resolution.

2. Weather/Meteorological Satellites:

   • INSAT (India): Geostationary. Sensors: Very High Resolution Radiometer (VHRR). Apps: weather forecasting, cyclone monitoring.

   • NOAA (USA): Polar-orbiting. AVHRR sensor. Apps: global weather, sea surface temp, vegetation index.

   • MetSat (India): Successor to INSAT series.

Indian Satellite Programme for RS

• IRS Series: Historical backbone. Started with IRS-1A (1988). Series includes LISS (Linear Imaging Self-Scanning), AWiFS (Advanced Wide Field Sensor) payloads. Applications: agriculture, water resources, urban planning.

• Chandrayaan-3 (2023):

  • Objectives: Demonstrate safe & soft landing on Moon, rover operations, in-situ scientific experiments.

  • Salient Features: Lander (Vikram), Rover (Pragyan), Propulsion module. Indigenous technology.

  • Payloads: LIBS (Laser Induced Breakdown Spectroscope), APXS (Alpha Particle X-ray Spectrometer), seismometer, etc. for lunar surface composition.

• Other Significant Missions:

  • Oceansat: Ocean color, sea surface temp, wind vectors (OCM, scatterometer).

  • Resourcesat: Advanced LISS & AWiFS for resources mapping.

  • Cartosat: High-resolution (sub-meter) stereoscopic imaging for cartography, DEM generation.

C. FUNDAMENTALS OF GEOGRAPHIC INFORMATION SYSTEMS (GIS)

• Definition: A computer-based system for capturing, storing, managing, analyzing, and displaying spatially referenced data.

• Core Objective: To support decision-making by integrating spatial (location) and attribute (descriptive) information.

• Key Components (6-Part Model):

  1. Hardware: Computer, storage, GPS, plotters.

  2. Software: GIS package (e.g., ArcGIS, QGIS), DBMS.

  3. Data: Most critical & costly component. Spatial & attribute data.

  4. People: Skilled users, managers, technicians.

  5. Procedures: Methods for data handling, analysis, and output.

  6. Network: For data sharing and distributed GIS.

GIS Data: Sources & Collection

• Primary Sources: Field surveys (GPS, total station), original maps, direct sensing.

• Secondary Sources: Published maps, census data, satellite imagery, existing digital databases.

• Objective of Data Collection: To create a comprehensive, accurate, and current spatial database that meets the specific needs of the GIS project (scale, accuracy, theme).

D. GIS DATA MODELS & STRUCTURES

Spatial Data Models

Attribute Data & Integration

• Attribute Table: Non-spatial database table linked to spatial features (via unique ID).

• Integration: One-to-one (e.g., polygon ↔ table row), one-to-many (e.g., point ↔ multiple photos), many-to-many relationships. The link between spatial and attribute data is the heart of GIS.

Coordinate Systems & Map Projections

• Geographic Coordinate System (GCS): Uses latitude/longitude on a spherical/ellipsoidal Earth (Angular units). No distortion but not suitable for measurement.

• Projected Coordinate System (PCS): Projects GCS onto a flat surface (planar). Uses linear units (meters, feet). All projections distort shape, area, distance, or direction.

• Types of Projections:

  • Cylindrical: Meridians & parallels intersect at right angles (e.g., UTM, Mercator). Good for equatorial regions.

  • Conical: Parallels are arcs, meridians are straight lines (e.g., Albers). Good for mid-latitude east-west extents.

  • Azimuthal: Direction from center is true. Good for polar regions or airline routes.

• Universal Transverse Mercator (UTM):

  • Concept: Global system of 60 zones (6° wide). Each zone has its own central meridian.

  • Characteristics: Conformal (preserves shape), uses metric system, minimizes distortion within zone. False Easting/Northing added to avoid negative coordinates.

  • Application: Large-scale topographic mapping, engineering projects.

E. GIS DATA INPUT, MANAGEMENT & CONVERSION

Data Input & Creation (Digitizing)

1. Manual (Tablet) Digitizing: Tracing map features on a digitizing tablet with a puck.

2. Automated (Scanning) Digitizing: Scanning map to raster, then heads-up digitizing (on-screen tracing using mouse) or automated vectorization.

3. Procedure for Shapefile Creation:

   • Set up coordinate system/projection.

   • Create new shapefile (point, line, polygon) with attribute fields.

   • Digitize features using appropriate tool.

   • Save and edit attribute table.

Georeferencing of Raster Images

• Purpose: Assign real-world coordinates to an image (e.g., scanned map, satellite image).

• Procedure:

  1. Load image & reference data (with known coordinates).

  2. Identify Ground Control Points (GCPs)—distinct, permanent features visible in both.

  3. Collect GCP coordinates (from reference data).

  4. Choose Transformation Model (e.g., Polynomial 1st order for linear, 2nd/3rd for non-linear rubber-sheeting).

  5. Apply transformation, resample (nearest neighbor, bilinear, cubic convolution), and save.

Data Conversion

• Raster to Vector (Vectorization): Converts grid cells to points/lines/polygons. Used for converting scanned maps. Issues: Thinning (line width), generalization, topology building.

• Vector to Raster (Gridding): Converts features to grid cells (e.g., for spatial modeling). Issues: Cell size selection (determines detail & file size), attribute assignment (e.g., polygon to cell value).

• Considerations: Loss of precision, increase in data volume, scale dependence, purpose of analysis (vector for discrete features, raster for continuous surfaces).

F. GIS SPATIAL ANALYSIS OPERATIONS

Query & Measurement

• Attribute Query: SELECT * FROM roads WHERE type = 'Highway'; (SQL logic).

• Spatial Query: "Select all wells within 500m of the river."

• Measurements: Distance (geodesic/planar), area, perimeter (automatically calculated for polygons).

Overlay Analysis

• Concept: Combining two or more thematic layers to create a new layer with combined attributes.

• Types (Polygon-on-Polygon):

  • Union: All areas from both layers (A OR B). Preserves all features.

  • Intersect: Only overlapping areas (A AND B). Creates new polygons from overlaps.

  • Identity: Features of input layer with attributes of overlay layer (like clip + identity).

  • Clip: Extract features of input layer within overlay layer boundary (A WITHIN B).

• Boolean Logic: AND (intersection), OR (union), XOR (symmetrical difference).

Neighborhood & Proximity Analysis

• Buffer Analysis: Creates zone(s) of specified distance around a feature.

  • Fixed Buffer: Constant distance (e.g., 100m from all rivers).

  • Dynamic Buffer: Distance varies by attribute (e.g., buffer based on traffic volume).

  • Applications: Noise pollution zones, pipeline/road influence areas, school service areas.

Classification & Reclassification

• Reclassification: Assigning new values to existing data classes (e.g., reclassify slope % into suitability classes: 0-5% = 1 (High), 5-15% = 2 (Medium)).

• Purpose: Simplify data, standardize classes, prepare for suitability modeling (e.g., weighted overlay).

G. INTEGRATION OF REMOTE SENSING & GIS

Role of RS as a Data Source for GIS

• Advantages:

  • Synoptic View: Captures large areas at once.

  • Multispectral: Provides spectral information beyond human vision.

  • Repetitive: Enables temporal monitoring (change detection).

  • Accessible: For remote/inaccessible areas.

• Pre-processing before GIS Integration:

  1. Geometric Correction: Remove distortions (systematic errors), georeference to map coordinates.

  2. Atmospheric Correction: Convert DN to radiance/reflectance for quantitative analysis.

Problems in Using RS Data in GIS

1. Geometric Distortions: Systematic (predictable, correctable: Earth curvature, sensor sweep) vs. Non-systematic (random: platform instability). Requires precise georeferencing.

2. Radiometric Errors: Sensor calibration drift, atmospheric effects → incorrect DN values. Requires atmospheric correction.

3. Resolution Mismatch: RS image resolution (e.g., 30m) may not match vector map scale (e.g., 1:10,000). Causes modifiable areal unit problem (MAUP). Requires resampling or generalization.

4. Data Format Compatibility: Different file formats (e.g., GeoTIFF, IMG, JPEG2000) and coordinate systems. Requires format conversion and projection transformation.

H. APPLICATIONS OF RS & GIS

Land Use/Land Cover (LULC) Change Detection

1. Acquire multi-temporal RS images (pre-processed).

2. Classify each image (supervised/unsupervised) into LULC classes.

3. Post-classification Comparison: Create change matrix (crosstabulation) showing transitions (e.g., Forest → Agriculture).

4. GIS Analysis: Overlay classified images, map change polygons, calculate area changes, identify hotspots.

5. Modeling: Use GIS to model drivers and predict future changes.

Water Resources Applications

• Surface Water: Mapping reservoirs/lakes (NDWI), wetland delineation, flood extent mapping (change detection), change in river course.

• Groundwater Potential: GIS overlay of drainage density, geology, lineaments (faults), slope, soil to identify recharge zones.

• Watershed Delineation: Using DEM (from RS stereo pairs or contour digitization) to derive flow direction, accumulation, and watershed boundaries.

• Snowmelt Runoff Estimation: Snow cover mapping (NIR/SWIR), melt rate modeling with temperature data.

• Drought Assessment: Vegetation indices (VCI, TCI) from RS combined with precipitation/temperature data in GIS.

Other Key Application Areas

• Urban Planning: Land use mapping, urban growth modeling (cellular automata), network analysis for traffic/emergency services, site suitability.

• Agriculture: Crop type mapping, crop health (NDVI), yield estimation, drought monitoring.

• Forestry: Deforestation mapping, forest fragmentation analysis, biomass estimation (LiDAR/Radar).

• Disaster Management: Landslide Susceptibility (slope, geology, rainfall), cyclone path tracking & impact assessment, earthquake vulnerability (fault lines, building density).

I. SYSTEMATIC vs. NON-SYSTEMATIC ERRORS

[!TIP] Exam Key: Systematic errors are deterministic (use models/corrections). Non-systematic are stochastic (use statistical filters like low-pass). In geometric correction, systematic errors are removed first (using sensor model), then non-systematic (using GCPs & polynomial fit).