Concept, Approach, Calibration, Atmospheric Compensation, Research and Applications
Imaging Spectroscopy is the acquisition of images where for each spatial resolution element in the image a spectrum of the energy arriving at the sensor is measured. These spectra are used to derive information based on the signature of the interaction of matter and energy expressed in the spectrum. This spectroscopic approach has been used in the laboratory and in astronomy for more than 100 years.
As an example, a reflectance spectrum of a mixture of three common rock forming minerals is shown below. Each mineral is a different molecule that absorbs energy in different regions of the spectrum. From the spectrum the three minerals can be identifed and with radiative transfer analysis the relative concentrations of the minerals can be determined.
Reflectance Spectrum of a Three Mineral Mixture
AVIRIS Spectral Sampling
The same spectrum is shown below as measured through the multispectral bands of the Landsat Thematic Mapper. The three molecules cannot be unambiguously identified, nor the relative concentrations determined.
Reflectance Spectrum of a Three Mineral Mixture
Landsat Thematic Mapper
The example above was for minerals found on the surface of the Earth. However, imaging spectroscopy is applicable wherever research, environmental and application questions can be posed in terms of molecules, scatterers and energy sources expressed in the solar reflected spectrum.
IMAGING SPECTROSCOPY IN THE SOLAR REFLECTED SPECTRUM
Quantitative measurement of components of the Earth System from calibrated spectra acquired as images for scientific research and applications.
Measure the upwelling radiance spectrum from 400 to 2500 nm at 10 nm resolution Based upon the molecular absorptions and constituent scattering characteristics expressed in the spectrum:
- Detect and identify the surface and atmospheric constituents present
- Assess and measure the expressed constituent concentrations
- Assign proportions to constituents in mixed spatial elements
- Delineate spatial distribution of the constituents
- Monitor changes in constituents through periodic data acquisitions
- Simulate, calibrate and intercompare sensors
- Validate, constrain and improve models
RESEARCH AND APPLICATIONS
Through measurement of the solar reflected spectrum, a range of scientific research and application is being pursed using signatures of energy, molecules and scatterers in the spectra measured by AVIRIS:
- Atmosphere: water vapor, clouds properties, aerosols, absorbing gases...
- Ecology: chlorophyll, leaf water, lignin, cellulose, pigments, structure, vegetation species and community maps, nonphotosynthetic constituents...
- Geology and soils: mineralogy, soil type...
- Coastal and Inland waters: chlorophyll, plankton, dissolved organics, sediments, bottom composition, bathymetry...
- Snow and Ice Hydrology: snow cover fraction, grainsize, impurities, melting...
- Biomass Burning: subpixel temperatures and extent, smoke, combustion products...
- Environmental hazards: contaminants directly and indirectly, geological substrate...
- Calibration: aircraft and satellite sensors, sensor simulation, standard validation..
- Modeling: radiative transfer model validation and constraint...
- Commercial: mineral exploration, agriculture and forest status...
- Algorithms: autonomous atmospheric correction, advanced spectra derivation...
- Other: human infrastructure...
With current technology, it has become possible to build sensors than can measure spectra as images of the Earth at high spectral and spatial resolution. Below is a depiction of how NASA's Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) measures spectra from 400 to 2500 nm at 10 nm intervals across the solar reflected spectrum. This chart shows the 224 AVIRIS spectral channels, a typical transmittance spectrum of the Earth's atmosphere and 6 of the multispectral bands measured by the Landsat Thematic Mapper.
AVIRIS Measurements in the Solar Reflected Spectrum
IMAGING SPECTROSCOPY PROJECTS
GENERALIZED IMAGING SPECTROSCOPY EXPERIMENT OUTLINE
1) A research, environmental or application question that is connected to the distribution of materials that interact with solar reflected energy.
- Vegetation (species type, chemistry, absorbing and scattering constituents, etc.) natural, managed forests, agriculture, etc.
- Coastal and inland water (in water constituents, bottom constituents, etc.)
- Human generated materials
- Hazards (contaminants, etc.)
- Atmosphere (absorbers, scatterers, etc.)
- Rocks and soils (minerals, altered minerals, etc.)
- Biomass burning (cellulose, leaf water, fire temperature, etc.)
Note: additional research, environmental or application questions may be pursued through correlations with materials that interacts with solar reflected energy.
2) Reflectance or equivalent signatures of materials of interest
- Spectral library of material signatures
- Physical model of material signatures
- Empirical strategy for deducing material signatures
3) Acquisition of imaging spectrometer data for optimal expression of solar reflected signatures.
- Clear sky and sun at optimal orientation
- Material of interest expressed (e.g. seasonal variation in vegetation)
3a) Contemporaneous validation data measurements
- If this is a validation experiment, then independent measurement of the distribution of the materials of interest will support validation of the results.
4) Have a strategy for atmospheric correction
4a) Measurements that support atmospheric correction.
- Reflectance measurements from one or two quasi homogeneous targets will generally improve the atmospheric correction
5) Algorithms and software that derive the expression of the material of interest from the atmospherically corrected reflectance imaging spectroscopy data set.
- Multiple endmember unmixing
- Empirical orthogonal function analysis
6) Algorithms that transform the resulting material distribution images to map coordinates.
7) Apply the derived information and address the research, environmental or application question.
8) Results and Reporting