Understanding HSI and MSI Bandsets

Overview

Aurora's spectral indices table often lists two formulas for the same index—one for Pixxel's Hyperspectral Imagery (HSI) and another for Sentinel-2 Multispectral Imagery (MSI). Although these indices may represent the same concept, their formulas are not always identical because they are built using data from different sensor technologies. Some indices are also exclusive to HSI, meaning they cannot be generated using Sentinel-2 imagery.

Hyperspectral Imagery (HSI) captures reflectance across many narrow, contiguous wavelength bands. By dividing the spectrum into dozens or even hundreds of thin slices, Pixxel's hyperspectral sensors can detect subtle spectral signatures that would otherwise be difficult to distinguish.

Multispectral Imagery (MSI), such as Sentinel-2, captures reflectance across a much smaller number of broader wavelength bands. These wider bands are sufficient for many standard vegetation and environmental indices but cannot resolve fine spectral features with the same precision as hyperspectral imagery.

A simple way to understand the difference is to compare two cameras. MSI captures the major colors and features of a scene, much like a standard camera. HSI captures hundreds of subtle spectral variations, like a specialized camera that reveals details invisible to the human eye. Both observe the same scene, but hyperspectral imagery provides a much richer level of information.

Why are formulas the different

The formulas differ because HSI and MSI sensors observe different portions of the electromagnetic spectrum.

HSI formulas reference Pixxel's narrow hyperspectral band IDs (for example, B011, B071, or B118), while MSI formulas use Sentinel-2's broader band IDs (such as B02, B03, B04, and B08).

Even when both sensors support the same spectral index, they must use the bands available on that sensor. Since the wavelength ranges, bandwidths, and center wavelengths differ, the resulting formulas—and often the numerical outputs—are naturally different.

Why some Indices only work with HSI

Some indices display NA under the Sentinel-2 (MSI) column. This does not indicate that Aurora is missing the calculation. Instead, it means Sentinel-2 does not capture the narrow wavelength region required to calculate that index.

Certain vegetation, mineral, and water signatures exist within extremely narrow portions of the spectrum. When a multispectral sensor uses a broad band that spans this region, the unique signal becomes averaged together with neighboring wavelengths, making the feature impossible to isolate. Pixxel's hyperspectral sensors overcome this limitation by measuring these narrow wavelength ranges individually.

In short, an index is HSI-only whenever it relies on spectral detail that cannot be resolved by the broader bands available in multispectral imagery.

Reading Band IDs

Aurora uses different band naming conventions depending on the imagery source.

• Pixxel HSI Bands — IDs such as B011, B071, and B118 represent individual narrow hyperspectral bands. Since Pixxel captures many bands, the numbering extends much higher than conventional multispectral sensors.

• Sentinel-2 MSI Bands — IDs such as B02, B03, B04, and B08 follow the standard Sentinel-2 naming convention, where each identifier corresponds to one of the satellite's broader multispectral bands.

Always verify the exact center wavelength and bandwidth of Pixxel HSI band IDs using Aurora's official band reference before quoting wavelength values to customers. Sentinel-2 band specifications are publicly documented by the European Space Agency (ESA), but should also be confirmed against the latest official documentation before being referenced externally.

Frequently Asked Questions

Why are there two formulas for the same index?

Aurora supports multiple imagery sources. Since Pixxel HSI and Sentinel-2 MSI capture different spectral bands, each sensor requires its own version of the formula that uses the bands available on that platform.

Why is the Sentinel-2 formula marked as NA for some indices?

Some indices depend on narrow spectral features that Sentinel-2's broader bands cannot isolate. These indices can only be generated from hyperspectral imagery because the required spectral detail is unavailable in multispectral data.

Does HSI replace MSI?

No. Both technologies serve different purposes. MSI provides extensive historical coverage and supports many widely used indices, making it suitable for numerous operational workflows. HSI complements MSI by enabling more detailed analysis and supporting advanced indices that require higher spectral resolution.

Which imagery should I use?

The appropriate choice depends on the customer's use case

• Use HSI when the required index is unavailable for Sentinel-2 or when maximum spectral precision is needed.
• Use MSI when historical archives, broader geographic coverage, or cost efficiency are the primary priorities and the required index is supported.
• For advanced analytical requirements where both options are available, HSI generally provides more detailed spectral information and finer analytical results.