NASA imaging spectrometer enhances airborne remote sensing

Imaging spectrometers used to observe the Earth through thermal infrared differentiate spectral signatures for features of the Earth's crust, fire fumes or plumes, and different plant types. The data collected is used in environmental monitoring and defense, and it is vital to earth science.

To identify spectral signatures on Earth, we typically need high resolution spectroscopy since there are usually mixtures of different spectra in every observation. Having more than one pure feature in a scene sometimes confuses the spectral classification algorithms, so more spectral channels enable greater accuracy. Existing spectrometers have a limited number of spectral bands, relatively small fields of view, and are bulky. We would like to capture high resolution spectra over large swaths in order to cover different spatial areas and to correlate data. A larger swath means less airplane time since a particular area can be covered more quickly. We hope eventually to put a system into space to explore Earth and other planets, so a small and lightweight structure would be advantageous.

Most of the earliest spectrometer designs would now be considered overly large with limited capabilities. Examples include the Spatially Enhanced Broadband Array Spectrograph System, a space-based spectrometer developed by Aerospace,¹ and the MODIS/ASTER airborne simulator, an airborne system developed by NASA.² Both have limited spectral capability and are resource-hungry. Aerospace recently demonstrated a semi-compact airborne thermal spectrometer,³ but it requires a scan mirror to build up an image and a power-hungry detector cooling system.

To get around these problems, we incorporated a sophisticated optical and mechanical system⁴ that uses a two-mirror, wide swath angle, snap-together telescope; a high efficiency concave diffraction grating; a Dyson relay optical system; and a quantum well infrared photodetector (QWIP).

We fabricated gratings with low scatter using electron-beam lithography techniques developed at NASA's Jet Propulsion Laboratory (JPL). These allow for gratings with several millimeters of height variation. We chose J. Dyson's⁵ very compact optical system, which includes a Seidel-corrected unit magnifier composed of a single lens and a concave mirror. It provides high optical throughput, and it limits the number of optical elements in the imaging spectrometer system, producing greater spatial and spectral accuracy. The spectrometer was mounted aboard aircraft for test flights (see Figure 1).

Figure 1. The lightweight spectrometer mounted aboard aircraft. Test flights were carried out in summer 2012.

QWIP technology uses the photoexcitation of electrons between the ground state and the first excited state in the conduction band quantum well. It has been used successfully in commercial hand-held field units for more than a decade, but this is the first integration of the QWIP in a spectrometer system for earth science studies that require accurately calibrated data.

Our imaging spectrometer can assemble improved datasets that enable a better understanding of satellite information. For example, it can assist in decoupling emissivity and temperature data, which is vital for understanding satellite information on Earth's ecosystems. During a remote measurement, wavelength-dependent emissivity is coupled to the local environmental temperatures, and this inherent mixing needs to be understood to fully appreciate the information content of the scene.⁶

We used two tests to verify the performance of our imaging spectrometer. For radiometric performance, we undertook a National Institute of Standards and Technology (NIST) traceable transfer calibration on our electro-optic blackbody to verify its performance between 5°C and 30°C. JPL has multiple NIST-traceable blackbodies with a stability at 25°C of +/−0.0007°C and a thermistor standard probe with an accuracy of 0.0015°C over 0–60°C and stability/year of 0.005. We ramped the blackbody from 5°C to 30°C and then left it to drift in 5°C increments to end up back at 5°C. We took frames at each interval to check for temporal artifacts as well as single frame noise equivalent temperature difference per spectral band, and to determine any spectral non-linearity. The results showed that the spectrometer's performance exceeded standard NIST expectations.

We ran the system outdoors and under direct sunlight to characterize its usefulness for remote sensing. Calculating radiance based on the data and plotting atmospheric water band absorption, we could compare with output from the Moderate Resolution Atmospheric Transmission (MODTRAN) program and Fourier transform imaging spectrometers, respectively. These comparisons verified the performance of our spectrometer and demonstrated its viability for earth science applications.

We successfully flew the system at low altitude for demonstration purposes in summer 2012 and plan further flights in early 2013.