Coastal Services Center

---

Hyperspectral Systems

---

Summary | Different Data Products | Applied Uses | Specifications | Data Ordering Details | Frequently Asked Questions | For More Information

Summary

Hyperspectral sensors are passive sensors that acquire simultaneous images in many relatively narrow, contiguous and/or non-contiguous spectral bands through the ultraviolet, visible and infrared portions of the electromagnetic spectrum. These systems have been traditionally used in laboratory settings but now scientists can acquire hyperspectral information via aircraft and satellite platforms. With this increased amount of spectral data, scientists have the ability to derive specific information about the earth's surface from an image pixel using diagnostic or characteristic absorption features in its spectral signature . The signature is generated by the information contained in the numerous spectral bands acquired by the sensor. The number of images collected by these sensors can be so great; the datasets are often referred to as "data cubes ."

Different Data Products

Point to the names below to view the different data products.

CASI
AVIRIS
AISA
Probe-1

Return to top

---

Applied Uses

Examples of how airborne hyperspectral data are being used:

Effects of Tidal Restriction in Marsh Habitats Structures such as roads, rail beds, dikes, and tide gates restrict tidal flow in coastal areas–which can alter the landscape and negatively impact the biodiversity and function of salt marshes. Coastal managers can use hyperspectral remote sensing data to determine the effects of tide restriction on marsh habitat and help prepare for future restoration and preservation projects.

Other Potential Uses:

• Oil Detection
• Vegetation Species Mapping
• Vegetation Stress Identification
• Water Quality Assessment
• Geologic Mapping

Return to top

---

Specifications

Spatial Resolution

In general, these technologies collect imagery with spatial resolutions ranging from 1 - 10 meters.

The following table lists specifications for a few of the existing hyperspectral systems:

Sensor	Acquisition Mode	Spectral Resolution	Swath Width	Radiometeric Resolution
AVIRIS	Whiskbroom Scanner	224 contiguous spectral channels over the wavelength range of 400 to 2500 nanometers	614 pixels	12 BIT
CASI 1500	Pushbroom Scanner	288 programmable channels over the wavelength range of 380 – 1050 nanometers	1500 pixels	14 BIT
AISA	Pushbroom Scanner	286 programmable channels over the wavelength range of 430 – 900 nanometers	360 pixels	14 BIT
Probe 1	Whiskbroom Scanner	128 spectral channels over the wavelength range of 400 – 2500 nanometers	512 pixels	11 BIT

Temporal Resolution

Revisit rate is dependent upon flight scheduling.

Return to top

---

Data Ordering Details

Data Acquisition	Many private companies, academic institutions, and government agencies produce and provide hyperspectral imagery.
Timing	Many time constraints are associated with hyperspectral imagery collection and delivery: Flight schedules can be delayed due to seasonal restrictions, weather, and environmental factors. Project areas may be large enough that multiple flights are needed. Post processing can be time consuming. Producing additional deliverables can delay the delivery schedule.
Cost	Cost can vary depending on size of project, spatial accuracy, and project location. Cost may also increase based on specific accuracy requirements, additional post-processing or licensing restrictions.
Data Formats/Software Needed	There are no standard file formats for hyperspectral imagery. Typically, vendors deliver imagery in the Generic Binary format.
Projections	Hyperspectral data can be delivered in many different projections and datums.
Licensing	Licensing restrictions vary for each vendor. Many providers do not have restrictions on their data products, but some companies do require licensing.

Return to top

---

Frequently Asked Questions

Data Aquisition

Can I download or request hyperspectral imagery on-line?
Several vendors have example imagery available at their Web sites.

Cost

How much does it cost?
Cost can vary depending on size of project, environmental constraints, spatial resolution, spectral resolution and level of post-processing.

Data Formats/Software Needed

What types of software are required?
Hyperspectral imagery is most commonly formatted as a Generic Binary file. Viewing, processing and analysis of these images will require specialized software, including:

• RSI ENVI^®
• ERDAS IMAGINE

Is freeware available for using these data?
A number of freeware viewers will allow you to view image data and perform simple tasks. It is recommended that you contact the software companies for more details. The following are some examples of freeware viewers:

• ENVI Freelook 3.1 by Research Systems, Inc.
• Geomatica® FreeView by PCI Geomatics

What are some of the hyperspectral imagery products available?

• Radiometrically calibrated radiance data (Radiance Units)
• Atmospherically corrected data (Reflectance Units)
• Spectral libraries
• Orthorectified images

General Questions

How do clouds and the atmosphere affect the data?
Airborne hyperspectral imagery can be collected during cloudy days by flying below the clouds. This works well with consistent overcast conditions as the ambient lighting on all frames is comparable. Cloud shadows can be problematic with these systems.

It is recommended that hyperspectral images be post processed for the influence of atmospheric constituents. Typically, hyperspectral images are delivered in radiance units. These images contain absorption features caused by Oxygen and Water in the atmosphere thus altering the spectral signature of surface features. Hyperspectral images must be converted to reflectance units if they are to be used with any laboratory or field collected spectra for analysis. This is usually done using a radiative transfer model such as ACORN which attempts to account for scattering and absorption which occurs at specific wavelengths. Most vendors perform this step for their clients.

What level of technical expertise is needed to use imagery?
A general understanding of the software is needed to view the imagery. Image processing and analyses of hyperspectral data may requires a highly trained technical person with a strong background in remote sensing.

Return to top

---

For more Information

• ENVI
• AVIRIS
• Probe-1
• NOAA Coastal Remote Sensing Frequently Asked Questions

Return to top