Spatial, Spectral, Temporal, and Radiometric Resolution

The four resolution tradeoffs that shape every remote-sensing product

Published

April 4, 2026

Before You Start

You should know
That a sensor has limits and that measurement always involves compromise.

You will learn
What the four major kinds of resolution mean and why improving one often comes at the expense of another.

Why this matters
Many remote-sensing mistakes come from asking a product to answer a question it was never designed to answer.

If this gets hard, focus on…
Spatial means where, spectral means which wavelengths, temporal means when, and radiometric means how finely the signal is measured.

A 10 m pixel can map field boundaries better than a 1 km pixel. A hyperspectral sensor can distinguish subtle absorption features better than a broad-band sensor. A daily revisit is better for rapidly changing events than a monthly revisit. A 12-bit detector can distinguish smaller signal differences than an 8-bit detector. None of these improvements is free. Sensor design is a tradeoff problem.

This chapter gives a compact language for that tradeoff space. Once you understand the four resolutions, sensor choice becomes much easier to reason about.

Resolution Tradeoffs

Every Sensor Buys One Kind Of Clarity By Giving Up Another

The four resolutions are easiest to remember when they are tied to different kinds of questions. Spatial asks how big each ground unit is, spectral asks how finely wavelength is sampled, temporal asks how often the scene is revisited, and radiometric asks how many distinct signal levels the detector can separate.

Spatial

How Big Is One Pixel?

coarse pixels fine pixels

Smaller pixels separate roads, channels, and field edges that coarse pixels merge into one mixed measurement.

Spectral

How Narrow Are The Bands?

wide bands blur detail narrow bands keep features

Broad bands average over absorption features; narrow bands preserve the chemistry or material contrast inside the curve.

Temporal

How Often Do We Return?

daily-ish fast events need dense revisits

Frequent return times capture floods, fire fronts, and crop phases that sparse revisit schedules can miss entirely.

Radiometric

How Many Levels Can We Separate?

few levels many levels

More radiometric levels make subtle brightness differences visible instead of rounding them into the same digital value.

One scene can be sampled with different priorities. A good product is not the one with the biggest numbers everywhere; it is the one whose tradeoffs fit the question.

1. The Question

Why can one remote-sensing product be excellent for crop monitoring but poor for detailed urban mapping, while another is excellent for terrain detail but poor for biochemical discrimination?

Usually the answer lies in resolution:

  • spatial: how large is each pixel?
  • spectral: how narrow and numerous are the wavelength bands?
  • temporal: how often do we get a new observation?
  • radiometric: how finely can the detector distinguish signal levels?

2. Spatial Resolution

Spatial resolution is the ground size represented by one pixel.

Examples:

  • 1 m pixel: good for small roads, rooftops, individual crowns
  • 10 m pixel: good for fields, streams, forest patches
  • 250 m pixel: good for regional vegetation patterns
  • 1 km pixel: good for continental climate signals

Worked Example By Hand

If a pixel is 30 m × 30 m, its area is:

A = 30 \times 30 = 900\ \text{m}^2

If a small pond covers only 200\ \text{m}^2, then it fills less than one full pixel:

\frac{200}{900} \approx 0.22

So the pond will likely appear as a mixed pixel rather than a clean water pixel.

3. Spectral Resolution

Spectral resolution describes how finely the sensor samples wavelength.

  • broad bands: few, wide wavelength intervals
  • narrow bands: many, thin wavelength intervals

Broad bands are often enough for:

  • simple land-cover separation
  • NDVI-style indices
  • basic change mapping

Narrow bands are better for:

  • mineral identification
  • pigment and water-content retrieval
  • subtle absorption-feature analysis

4. Temporal Resolution

Temporal resolution is how often the sensor returns to observe the same place.

This matters because some processes change:

  • slowly, like glacier retreat
  • seasonally, like vegetation growth
  • rapidly, like flooding, fire, or storm damage

A product with coarse spatial detail but daily observations can be more useful for drought monitoring than a fine-resolution product that returns only rarely.

5. Radiometric Resolution

Radiometric resolution is how finely the sensor distinguishes signal intensity.

It is often described in bits:

  • 8-bit sensor: 2^8 = 256 levels
  • 12-bit sensor: 2^{12} = 4096 levels
  • 16-bit sensor: 2^{16} = 65536 levels

Higher radiometric resolution helps when the surface differences are subtle and you need the detector to notice small changes in radiance or reflectance.

6. The Tradeoff Picture

The point is not that one sensor type is best. The point is that they are optimized for different questions.

7. A Practical Reading Habit

When you encounter a remote-sensing product, ask:

  1. What is the pixel size?
  2. How many bands does it have, and how wide are they?
  3. How often can the place be observed?
  4. How finely can the detector distinguish signal levels?

Those four questions often tell you more about the usefulness of the product than its brand name does.

8. If This Gets Hard, Focus On

  • spatial = pixel size
  • spectral = wavelength detail
  • temporal = revisit timing
  • radiometric = signal sensitivity
  • good sensing is always a tradeoff, not a perfect-all-at-once design

That is enough to make later chapters on multispectral, thermal, hyperspectral, LiDAR, and radar much easier to place in context.