What Is Remote Sensing?
Measuring Earth from a distance with light, heat, sound, and microwave energy
Before You Start
You should know
That maps can be built from measured data and that light interacts with surfaces in different ways.
You will learn
What remote sensing actually measures, how passive and active systems differ, and why a sensor never sees “the thing itself” directly.
Why this matters
Most modern computational geography depends on Earth observation. Before using an index, a classification, or a satellite product, you need a model of how the measurement was made.
If this gets hard, focus on…
The core idea that a sensor records energy, and we infer surface properties from that energy.
When people first hear “remote sensing,” they often imagine a camera in space taking a picture. Sometimes that is close enough. Often it is not. A thermal sensor does not record visible colour; it records emitted radiation. Radar does not wait for sunlight; it transmits its own microwave pulse and measures the backscatter. LiDAR does something similar with laser light. GRACE does not image water at all; it infers mass redistribution from gravity change. In every case, the measurement reaches the analyst only after physics has transformed the surface into a signal.
That is why remote sensing belongs in a modelling book. The sensor is not a magical eye. It is part of a chain: energy interacts with the atmosphere, then the surface, then the sensor, then the data system that calibrates and packages the result. This chapter introduces that chain and the basic language needed to think clearly about Earth observation before we move into more specialized chapters.
1. The Question
How can we learn something about Earth’s surface without touching it directly?
The answer is: by measuring energy that has interacted with the surface.
Sometimes that energy comes from the Sun. Sometimes it is emitted by the surface itself. Sometimes the sensor provides the energy and measures the return.
So remote sensing always asks two linked questions:
- What energy is being measured?
- How did the surface and atmosphere modify that energy before the sensor recorded it?
2. The Sensing Chain
Remote sensing works through a chain of transformations:
- Energy source
- Atmosphere
- Surface interaction
- Sensor measurement
- Data product and interpretation
Sunlight, thermal emission, laser pulse, radar pulse, or acoustic pulse.
Scattering, absorption, and emission change the signal before and after surface contact.
The surface reflects, emits, absorbs, or scatters energy depending on material and geometry.
The instrument samples the signal with finite spatial, spectral, and radiometric limits.
Calibration and modelling turn raw measurement into maps, indices, temperatures, heights, or classifications.
The important habit is to resist jumping straight to the product. A vegetation index, a temperature map, or a flood mask is never the first thing measured.
What A Sensor Actually Records
Most remote-sensing instruments record one of four broad kinds of quantity:
- reflected radiation
- emitted radiation
- backscattered energy
- travel time or phase
Those measurements can then be converted into things we care about:
- vegetation greenness
- land surface temperature
- flood extent
- forest canopy height
- soil moisture
- ice loss
That conversion step is the model.
This is the main habit the chapter is trying to build. A remote-sensing product is never the first thing measured. Energy is generated, altered by the atmosphere, modified by the surface, sampled by a sensor, and only then converted into a product we can interpret.
3. Passive And Active Remote Sensing
Passive Systems
A passive sensor does not provide its own illumination. It records energy already present in the environment.
Examples:
- visible and near-infrared satellite imagery
- thermal infrared imaging
- passive microwave sensing
Passive systems depend on what the environment provides:
- reflected solar radiation for optical sensing
- emitted thermal radiation for thermal sensing
Active Systems
An active sensor sends out energy and measures the return.
Examples:
- radar
- LiDAR
- sonar
Active systems are powerful because they control the signal source. That often means:
- day and night operation
- better control over ranging
- some ability to penetrate cloud, canopy, or water depending on wavelength
This is a teaching chart, not a strict rulebook. Thermal infrared is passive but can still work at night because the surface emits its own radiation. The bigger lesson is that “remote sensing” is not one technology.
4. Pixels, Footprints, And Mixed Surfaces
A sensor does not usually observe an infinitesimal point. It observes a finite footprint on the ground.
For an imaging sensor, that footprint becomes a pixel in the image.
If the pixel contains:
- 100% forest, interpretation is easy
- 50% forest and 50% soil, interpretation is harder
- forest, water, and road together, it becomes a mixing problem
That is why even a beautiful image is not direct truth. Each pixel is already a summary over space.
Worked Example By Hand
Suppose a pixel is 30 m × 30 m, so its ground area is:
A = 30 \times 30 = 900\ \text{m}^2
If 60% of that pixel is vegetation and 40% is bare soil, then:
- vegetation area = 0.60 \times 900 = 540\ \text{m}^2
- bare soil area = 0.40 \times 900 = 360\ \text{m}^2
The sensor does not record two separate numbers. It records one number per band that mixes the signal from both parts.
That is the foundation of spectral mixing, classification uncertainty, and sub-pixel modelling.
5. Why Geometry Matters
The same field can look different depending on:
- sun angle
- sensor view angle
- surface slope and aspect
- shadowing
- canopy structure
So remote sensing is not only about material. It is also about geometry.
That is why later chapters on solar geometry, satellite overpasses, and canopy reflectance matter. The signal is shaped by where the source, surface, and sensor sit relative to one another.
6. If This Gets Hard, Focus On
- the sensor measures energy, not “the phenomenon itself”
- passive sensors record available energy; active sensors send out their own
- a pixel is a finite footprint, not a perfect point
- atmosphere and geometry both alter the signal
- the data product is always one modelling step beyond the raw measurement
That foundation is enough to make the rest of the remote-sensing sequence feel much less mysterious.