The ultimate remote sensing benefits of the high resolution Infrared radiance spectrometers will be realized with
their geostationary satellite implementation in the form of imaging spectrometers. This will enable dynamic
features of the atmosphere's thermodynamic fields and pollutant and greenhouse gas constituents to be observed
for revolutionary improvements in weather forecasts and more accurate air quality and climate predictions.
As an important step toward realizing this application objective, the Geostationary Imaging Fourier Transform
Spectrometer (GIFTS) Engineering Demonstration Unit (EDU) was successfully developed under the NASA New
Millennium Program, 2000-2006. The GIFTS-EDU instrument employs three focal plane arrays (FPAs), which
gather measurements across the long-wave IR (LWIR), short/mid-wave IR (SMWIR), and visible spectral bands.
The raw GIFTS interferogram measurements are radiometrically and spectrally calibrated to produce radiance
spectra, which are further processed to obtain atmospheric profiles via retrieval algorithms. The radiometric
calibration is achieved using internal blackbody calibration references at ambient (260 K) and hot (286 K)
temperatures. The absolute radiometric performance of the instrument is affected by several factors including
the FPA off-axis effect, detector/readout electronics induced nonlinearity distortions, and fore-optics offsets.
The GIFTS-EDU, being the very first imaging spectrometer to use ultra-high speed electronics to readout
its large area format focal plane array detectors, operating at wavelengths as large as 15 microns, possessed
non-linearity's not easily removable in the initial calibration process. In this paper, we introduce a refined
calibration technique that utilizes Principle Component (PC) analysis to compensate for instrument distortions
and artifacts remaining after the initial radiometric calibration process, thus, further enhance the absolute
calibration accuracy. This method is applied to data collected during an atmospheric measurement experiment
with the GIFTS, together with simultaneous observations by the accurately calibrated AERI (Atmospheric
Emitted Radiance Interferometer), both simultaneously zenith viewing the sky through the same external scene
mirror at ten-minute intervals throughout a cloudless day at Logan Utah on September 13, 2006. The PC vectors
of the calibrated radiance spectra are defined from the AERI observations and regression matrices relating the
initial GIFTS radiance PC scores to the AERI radiance PC scores are calculated using the least squares inverse
method. A new set of accurately calibrated GIFTS radiances are produced using the first four PC scores in
the regression model. Temperature and moisture profiles retrieved from the PC-calibrated GIFTS radiances are
verified against radiosonde measurements collected throughout the GIFTS sky measurement period.