What is Multispectral Imagery?

Multispectral satellite imagery is one of the most widely used forms of Earth observation data. It underpins a large proportion of operational satellite workflows across government, mining, environmental monitoring, agriculture and disaster response.

In simple terms, multispectral imagery captures Earth data in multiple specific wavelengths of light - beyond what the human eye can see - allowing us to analyse vegetation, water, soil and built environments in far greater detail than standard imagery.

Seeing Beyond Natural Colour

A standard image captures three bands of light: red, green and blue (RGB). Multispectral sensors extend this by capturing additional parts of the electromagnetic spectrum, typically including:

Blue
Green
Red
Near-infrared (NIR)
Shortwave infrared (SWIR)
In some sensors, red-edge bands

Each band captures reflectance at a different wavelength and each wavelength responds differently depending on the material on the ground.

Sentinel-2 natural colour (RGB) image of Texas Downs, Western Australia — Texas Downs, Western Australia - the same scene rendered in natural colour (left) and colour-infrared (right). Vegetation along watercourses, almost invisible in RGB, lights up red in CIR. Sentinel-2 imagery (ESA) - processed by Terrabit.

Sentinel-2 colour-infrared (CIR) composite of the same area — Texas Downs, Western Australia - the same scene rendered in natural colour (left) and colour-infrared (right). Vegetation along watercourses, almost invisible in RGB, lights up red in CIR. Sentinel-2 imagery (ESA) - processed by Terrabit.

Rather than capturing what something looks like, multispectral imagery helps determine what something is made of and how it is changing over time.

How Multispectral Sensors Work

Multispectral satellites do not capture a single photograph in the traditional sense.

Instead, they collect multiple layers of data, each representing a specific wavelength range. These are known as spectral bands.

Each pixel in a multispectral image contains reflectance values across those bands. This allows analysts to:

Compare spectral responses between materials
Detect changes over time
Classify land cover types
Extract environmental indicators

Each band effectively acts as a different filter on the same location.

Why Different Materials Respond Differently

Every surface on Earth reflects and absorbs light differently depending on its composition.

Vegetation

Healthy vegetation reflects strongly in the near-infrared band. This response is the basis for vegetation health monitoring and indices such as NDVI.

Water

Water absorbs most infrared light, making it appear dark in NIR and SWIR bands. This characteristic is used for flood mapping and water body delineation.

Urban areas

Built environments such as concrete, asphalt and rooftops each have distinct spectral signatures that allow classification of urban land use and change detection.

Minerals and soils

SWIR bands are particularly useful in distinguishing soil types and mineral composition, making them valuable in exploration and environmental applications.

WorldView-2 natural colour image of arid landscape near Mt Kinahan, WA — Mt Kinahan, Western Australia - RGB versus NDWI. Water absorbs strongly in NIR and SWIR, so the Normalized Difference Water Index pulls subtle drainage and moisture features forward that natural colour cannot resolve. Imagery courtesy of Vantor (WorldView-2) - processed by Terrabit.

WorldView-2 NDWI water index of the same area highlighting watercourses — Mt Kinahan, Western Australia - RGB versus NDWI. Water absorbs strongly in NIR and SWIR, so the Normalized Difference Water Index pulls subtle drainage and moisture features forward that natural colour cannot resolve. Imagery courtesy of Vantor (WorldView-2) - processed by Terrabit.

This separation of materials is what makes multispectral data operationally valuable.

WorldView Legion natural colour image of agricultural land in Tweed, NSW — Tweed, New South Wales - RGB versus NDVI. The same scene re-expressed as a Normalized Difference Vegetation Index, where bright green indicates dense, actively photosynthesising canopy. Imagery courtesy of Vantor (WorldView Legion) - processed by Terrabit.

NDVI vegetation index of the same agricultural area — Tweed, New South Wales - RGB versus NDVI. The same scene re-expressed as a Normalized Difference Vegetation Index, where bright green indicates dense, actively photosynthesising canopy. Imagery courtesy of Vantor (WorldView Legion) - processed by Terrabit.

Multispectral vs Hyperspectral

Multispectral and hyperspectral imagery are often grouped together, but they differ significantly in spectral detail.

Feature	Multispectral	Hyperspectral
Number of bands	Typically 4-16	100+ narrow bands
Spectral detail	Moderate	Very high
Data volume	Lower	High
Processing complexity	Moderate	High
Typical use cases	Mapping, vegetation, water, land cover	Material identification, mineralogy

Multispectral imagery is the operational standard across most commercial Earth observation systems due to its balance of detail, cost and scalability.

Hyperspectral natural colour image of geological terrain in the Northern Territory — Northern Territory, Australia - the same hyperspectral capture in natural colour (left) and as a PCA-derived composite (right). With hundreds of narrow bands, hyperspectral data exposes material differences that even a rich multispectral scene cannot resolve. Imagery courtesy of Wyvern - processed by Terrabit.

Principal Component Analysis of the same hyperspectral capture — Northern Territory, Australia - the same hyperspectral capture in natural colour (left) and as a PCA-derived composite (right). With hundreds of narrow bands, hyperspectral data exposes material differences that even a rich multispectral scene cannot resolve. Imagery courtesy of Wyvern - processed by Terrabit.

Common Band Combinations

One of the most widely used applications of multispectral imagery is band combination analysis.

By assigning different spectral bands to RGB channels, it is possible to highlight features not visible in natural colour imagery.

Common examples include:

False colour infrared for vegetation analysis
Urban composites for built-up mapping
SWIR combinations for burn scars and soil moisture
Water-focused composites for flood extent mapping

These combinations are widely used in environmental monitoring, emergency response and resource planning workflows.

Normalized Burn Ratio (NBR) composite derived from Sentinel-2 — A Normalized Burn Ratio (NBR) composite - a SWIR-based combination used to map fire severity and post-fire recovery. Sentinel-2 imagery (ESA) - processed by Terrabit. For a deeper look, read our guide on band combinations and indices.

Key Applications

Multispectral imagery is used across a wide range of operational sectors:

Agriculture

Crop health monitoring
Drought assessment
Yield estimation

Environmental monitoring

Deforestation tracking
Vegetation Classifications
Water quality assessment

Emergency management

Flood mapping
Bushfire burn severity assessment
Cyclone and storm impact analysis

Mining and resources

Exploration targeting
Land disturbance monitoring
Rehabilitation tracking

Urban and infrastructure

Urban expansion monitoring
Construction activity tracking
Land use classification

Why Multispectral Imagery Is So Widely Used

Multispectral data remains the backbone of operational Earth observation because it offers:

Broad geographic coverage with frequent revisit
Sufficient spectral detail for environmental and land analysis
Scalable data volumes suitable for large-area monitoring
Established analytical methods and indices

It provides a practical balance between simplicity and analytical power.

Limitations

While highly effective, multispectral imagery has a number of operational constraints that should be considered when designing workflows.

Limited spectral bands compared to hyperspectral systems
Multispectral sensors capture a smaller number of broader wavelength bands, which limits the level of material discrimination possible compared to hyperspectral imagery.
Dependent on cloud-free optical acquisition
As an optical dataset, imagery quality and availability can be impacted by cloud cover, smoke, haze and atmospheric conditions.
Requires correct band selection and processing
Reliable outputs depend on appropriate band combinations, calibration and processing workflows. Incorrect configurations can lead to misleading interpretations.
Interpretation relies on domain expertise
Spectral responses are not always intuitive and require experience in remote sensing and environmental context to interpret accurately.

Where It Fits in Earth Observation

Multispectral imagery sits at the core of most Earth observation workflows.

It bridges the gap between:

Standard RGB satellite imagery used for visual interpretation
Hyperspectral data used for material-level analysis
Radar systems used for structural and all-weather monitoring

In most operational programs, multispectral data acts as the primary analytical layer for environmental and land-based decision-making.

Final Thoughts

Multispectral satellite imagery is one of the most important and widely used datasets in Earth observation.

It translates reflected light into structured information about the Earth's surface - enabling consistent monitoring of vegetation, water, land use and environmental change at scale.

As satellite constellations continue to expand and revisit times improve, multispectral data will remain central to operational decision-making across government, industry and environmental sectors.

Talk to a specialist →