Skip to main content

Skill Guide

Remote sensing physics

Remote sensing physics is the study of the interaction between electromagnetic radiation and Earth's surface materials and atmosphere, providing the physical basis for interpreting and extracting information from remotely sensed data.

This skill is foundational for transforming raw satellite or airborne imagery into actionable intelligence, enabling organizations to make data-driven decisions in agriculture, environmental monitoring, urban planning, and defense. Mastery directly impacts the accuracy, reliability, and cost-effectiveness of geospatial products and services.
1 Careers
1 Categories
9.0 Avg Demand
15% Avg AI Risk

How to Learn Remote sensing physics

1. **Fundamental Electromagnetic Theory**: Understand the properties of the electromagnetic spectrum, Planck's law, Stefan-Boltzmann law, and Wien's displacement law. 2. **Radiometric Quantities**: Grasp the definitions and units of radiance, irradiance, reflectance, and brightness temperature. 3. **Interaction Mechanisms**: Learn the core processes of absorption, scattering (Rayleigh, Mie), emission, and transmission as they relate to atmospheric windows and surface materials.
1. **Spectral Signature Analysis**: Move from theory to practice by analyzing the spectral signatures (reflectance vs. wavelength curves) of common materials like vegetation, soil, water, and urban structures using libraries like USGS or ECOSTRESS. 2. **Atmospheric Correction**: Apply and validate basic atmospheric correction algorithms (e.g., DOS, 6S) on real imagery, understanding the sources of error like aerosol type and water vapor. 3. **Sensor-Specific Physics**: Delve into the physics of specific sensor types (e.g., multispectral vs. hyperspectral, passive vs. active radar/LiDAR), focusing on their unique radiometric and geometric characteristics. Common mistake: Ignoring the BRDF (Bidirectional Reflectance Distribution Function) effects in heterogeneous landscapes.
1. **System-Level Design & Optimization**: Contribute to or lead the design of sensor systems by modeling signal-to-noise ratios (SNR), radiometric resolution, and swath-width trade-offs for specific application requirements. 2. **Advanced Inversion Modeling**: Develop or critically assess complex physical models for inverting surface parameters (e.g., soil moisture, leaf area index, chlorophyll content) from multi-angle or multi-temporal data, accounting for coupled land-atmosphere processes. 3. **Strategic Alignment & Mentoring**: Align remote sensing physics knowledge with organizational R&D goals, evaluate new sensor technologies for their physical suitability to business problems, and mentor junior scientists in rigorous physical interpretation.

Practice Projects

Beginner
Project

Spectral Signature Identification and Material Classification

Scenario

You are given a set of hyperspectral image cubes from a drone survey of a mixed agricultural and suburban area. Your task is to identify distinct land cover materials based on their spectral signatures.

How to Execute
1. **Data Preparation**: Load the hyperspectral data into ENVI or Python (using libraries like `spectral` or `rasterio`). Extract average spectra from visually distinct regions (e.g., a wheat field, an asphalt road, a lake). 2. **Reference Comparison**: Compare these extracted spectra to standard spectral libraries (USGS, ECOSTRESS). 3. **Signature Plotting**: Plot the reflectance vs. wavelength for each material and annotate key absorption features (e.g., chlorophyll at 0.68 µm, water at 1.45 µm). 4. **Simple Classification**: Use a minimum distance or spectral angle mapper (SAM) classifier to produce a preliminary land cover map based on the reference signatures.
Intermediate
Project

Atmospheric Correction Validation for NDVI Analysis

Scenario

A client provides a time series of Landsat-8 Top-of-Atmosphere (TOA) reflectance images for monitoring crop health via NDVI. They suspect atmospheric effects are causing temporal noise. Your job is to apply and validate atmospheric correction.

How to Execute
1. **Pseudo-Invariant Feature Selection**: Identify pseudo-invariant features (e.g., large concrete pads, deep water bodies) across the image series. 2. **Atmospheric Correction Application**: Apply a physics-based atmospheric correction model (like 6S or ACOLITE) to convert TOA to Bottom-of-Atmosphere (BOA) reflectance. 3. **Temporal Analysis**: Calculate NDVI for both TOA and BOA corrected images for the invariant features. 4. **Validation**: Quantify the reduction in temporal variance (standard deviation) of NDVI for the invariant features post-correction. Present the findings with plots showing the 'cleaning' of the temporal signal.
Advanced
Project

Fusion of SAR and Optical Data for Soil Moisture Retrieval

Scenario

You are tasked with developing a robust soil moisture monitoring service for a drought-prone agricultural region. Optical data is often hindered by cloud cover, while SAR (Synthetic Aperture Radar) penetrates clouds but is sensitive to surface roughness and vegetation.

How to Execute
1. **Physical Model Selection**: Choose a physically-based inversion model that couples radar backscatter (σ⁰) with optical vegetation indices (e.g., Water Cloud Model or Integral Equation Model). 2. **Data Preparation**: Procure and preprocess co-registered Sentinel-1 (SAR) and Sentinel-2 (optical) data. 3. **Model Parameterization & Calibration**: Use in-situ soil moisture measurements and vegetation surveys to calibrate model parameters (e.g., vegetation structure parameter, surface roughness). 4. **Validation & Uncertainty Quantification**: Validate the fused soil moisture product against ground station networks. Perform sensitivity analysis to identify which input parameters (vegetation water content, roughness) most influence retrieval uncertainty, and document the model's physical limitations.

Tools & Frameworks

Software & Platforms

ENVI/IDLQGIS + Semi-Automatic Classification Plugin (SCP)Python (Rasterio, GDAL, Py6S, sckit-learn)SNAP (Sentinel Application Platform)

ENVI is the industry standard for advanced spectral analysis and atmospheric correction. QGIS with SCP provides an accessible open-source workflow. Python libraries are essential for custom modeling, automation, and integration into data pipelines. SNAP is critical for processing ESA satellite data (Sentinel-1,2,3) with scientifically validated algorithms.

Data & Reference Libraries

USGS Spectral LibraryECOSTRESS Spectral LibraryASTER Spectral LibraryMODIS/VIIRS Land Product Subsets

These libraries provide ground-truth spectral signatures for materials, essential for calibration, validation, and training classification algorithms. MODIS/VIIRS subsets offer ready-to-use, well-calibrated global products for validation or as ancillary data.

Physical Models & Frameworks

6S (Second Simulation of the Satellite Signal in the Solar Spectrum)PROSAIL (Coupled PROSPECT leaf + SAIL canopy models)Water Cloud Model

6S is a foundational radiative transfer model for atmospheric correction. PROSAIL is the standard for simulating vegetation reflectance from biophysical parameters. The Water Cloud Model is a simple but effective empirical-physical model for relating SAR backscatter to vegetation and soil moisture. These are the building blocks for developing custom inversion algorithms.

Interview Questions

Answer Strategy

This tests the candidate's ability to connect fundamental electromagnetic interaction mechanisms across different parts of the spectrum. The answer should explicitly contrast the physical processes (selective absorption vs. dielectric properties and scattering). **Sample Answer**: 'In the optical spectrum, healthy vegetation reflects green light (0.5-0.55 µm) while strongly absorbing red and blue light for photosynthesis, giving it a green hue. In SAR, the interaction is governed by the dielectric properties of water and the physical structure of the canopy. At lower frequencies (e.g., L-band), waves penetrate the canopy and scatter off larger woody stems and the soil beneath, providing moisture information. At higher frequencies (e.g., X-band), waves scatter off the smaller leaves and branches near the top of the canopy, making it appear bright if the canopy is dense and wet, as the water content increases the dielectric contrast.'

Answer Strategy

This is a scenario-based question testing systematic problem-solving, knowledge of confounding factors, and application of physics to real-world anomalies. The core competency is diagnosing data discrepancies. **Sample Answer**: 'I would follow a physics-first diagnostic. 1. **Sensor & Metadata Check**: Verify for sensor calibration errors, viewing/illumination geometry differences (BRDF effects), and shadow from terrain. 2. **Atmospheric Influence**: Assess if a persistent, localized atmospheric anomaly (e.g., thin cirrus, aerosol plume) is differentially affecting the two fields in the red band, artificially lowering NDVI. 3. **Surface Biophysical Differences**: If those are clear, I'd examine higher-resolution imagery for subtle but critical differences in soil background (brighter soil increases the red reflectance, depressing NDVI), row orientation (changing shadow fraction), or early growth stage differences. The key is to isolate the variable by using additional spectral indices (e.g., SAVI for soil-adjusted vegetation) and atmospheric correction outputs to rule out non-vegetation signals.'

Careers That Require Remote sensing physics

1 career found