Series 5: Advanced Remote Sensing

Series Overview

Series 5: Advanced Remote Sensing

Series Overview

Remote sensing extends human perception across wavelengths invisible to the eye and provides synoptic Earth observation impossible from the ground. This series develops the physics and signal processing underlying modern Earth observation, from hyperspectral imaging through synthetic aperture radar to gravity field measurements.

Pedagogical Philosophy

Physics determines capability. Each wavelength interacts differently with Earth’s surface. Understanding electromagnetic and gravitational physics reveals what can be measured and what cannot.

Signal processing essential. Raw sensor data requires substantial processing. Students learn algorithms, not just interpretation.

Active vs passive dichotomy. Passive sensors (optical, thermal, microwave radiometry) detect natural emissions. Active sensors (LiDAR, SAR) illuminate targets. Different physics, different information.

Learning Objectives

  1. Interpret hyperspectral signatures and apply spectral unmixing
  2. Calculate surface temperature from thermal infrared with emissivity correction
  3. Process LiDAR point clouds for terrain and vegetation structure
  4. Implement SAR image formation and interferometry
  5. Analyze gravity field changes from GRACE satellite measurements
  6. Apply passive microwave for soil moisture and sea ice
  7. Select appropriate sensors for measurement objectives
  8. Quantify retrieval uncertainty from sensor characteristics

Model Sequence

Cluster S: Advanced Optical & Thermal (Models 52-53)

Model 52: Hyperspectral Imaging and Spectral Unmixing Continuous spectra. Absorption features. Linear unmixing. Spectral Angle Mapper. Mineral identification.

Model 53: Thermal Infrared Remote Sensing Brightness temperature. Emissivity correction. Split-window algorithms. Land surface temperature. Urban heat islands.

Cluster T: Active Ranging Systems (Models 54-57)

Model 54: Terrestrial LiDAR Point Clouds Range equation. Multiple returns. Ground classification. Canopy height models. Biomass estimation.

Model 55: Bathymetric LiDAR and Sonar Mapping Water penetration. Refraction correction. Multibeam sonar. Backscatter classification. Seafloor mapping.

Model 56: SAR Fundamentals and Applications Synthetic aperture. Range-azimuth resolution. Backscatter mechanisms. Polarimetry. Flood/ship detection.

Model 57: InSAR Deformation Monitoring Interferometric phase. Unwrapping. Subsidence measurement. Earthquake/volcano deformation. Coherence.

Cluster U: Geophysical Remote Sensing (Models 58-59)

Model 58: Gravity Remote Sensing (GRACE) Satellite gravimetry. Mass change detection. Groundwater depletion. Ice sheet mass balance. Geoid.

Model 59: Passive Microwave Radiometry Brightness temperature. Soil moisture retrieval. Sea ice concentration. Snow water equivalent. All-weather capability.

Mathematical Progression

Optical: Radiative transfer, linear algebra (unmixing)

Thermal: Stefan-Boltzmann, Planck’s law, atmospheric correction

LiDAR: Geometry, waveform processing, classification algorithms

SAR: Fourier processing, complex numbers, phase unwrapping

Gravity: Spherical harmonics, potential theory

Microwave: Dielectric properties, emissivity models

Computational Skills

  • Spectral signature analysis
  • Point cloud processing
  • SAR focusing algorithms
  • Phase unwrapping (branch cuts, minimum cost flow)
  • Spherical harmonic decomposition
  • Bayesian retrieval inversions

Prerequisites

Required: Series 1 (calculus, complex numbers)

Helpful: Series 2 Model 13 (radiation basics), Series 3 Model 29-30 (raster processing)

Not required: Prior remote sensing background

Entry Points

LiDAR applications: Models 54-55 accessible with Series 1

SAR/InSAR: Models 56-57 most mathematically demanding

Climate/hydrology: Models 58-59 (gravity, passive microwave)

Optical background: Models 52-53 extend visible/NIR knowledge

Key Insights

  1. Wavelength selection critical. Different wavelengths penetrate atmosphere, vegetation, water differently.

  2. Active sensors all-weather. Radar/LiDAR work through clouds; optical does not.

  3. Phase information powerful. InSAR detects millimeter deformations from satellite altitude.

  4. Resolution tradeoffs. Spatial vs spectral vs temporal. Cannot maximize all simultaneously.

  5. Atmospheric correction essential. Water vapor, aerosols corrupt surface signal.

  6. Viewing geometry matters. Incidence angle affects backscatter, thermal emission.

  7. Validation requires ground truth. Remote sensing retrievals are models, not direct measurements.

Applications

  • Mineral exploration (hyperspectral)
  • Urban heat mapping (thermal IR)
  • Forest structure (LiDAR)
  • Seafloor mapping (bathymetric LiDAR)
  • Flood extent (SAR)
  • Earthquake deformation (InSAR)
  • Groundwater monitoring (GRACE)
  • Soil moisture (passive microwave)

Sensor Platforms Covered

Optical/Thermal: Landsat, MODIS, ASTER, ECOSTRESS

Hyperspectral: AVIRIS, EnMAP, PRISMA

LiDAR: Airborne systems, ICESat-2, GEDI

SAR: Sentinel-1, ALOS-2, RADARSAT, TerraSAR-X

Gravity: GRACE, GRACE-FO

Microwave: SMAP, SMOS, AMSR2

Extensions

For classification: Machine learning on multispectral/hyperspectral (future)

For time series: Change detection methods from Series 3 Model 41

For validation: Ground-based measurements from Series 2

Estimated Time

Per model: 4-5 hours (physics + processing + application)

Full series: 30-40 hours

SAR sequence (56-57): 10-12 hours (most complex)

Prerequisites: Series 1 required. Selected Series 2 models helpful for radiation physics background.