Engineering topics
Publications and source records attributed to Pagnutti, Mary.
The objective of this project is to perform radiometric vicarious calibrations of IKONOS satellite imagery and compare with Space Imaging calibration coefficients. This viewgraph presentation describes characterization undertaken at Lunar Lake Playa, Nevada; and Stennis Space Center, Mississippi.
Scientists at NASA's Earth Science Applications Directorate are creating a well-characterized Verification & Validation (V&V) site at the Stennis Space Center. This site enables the in-flight characterization of remote sensing systems and the data they acquire. The data are predominantly acquired by commercial, high spatial resolution satellite systems, such as IKONOS and QuickBird 2, and airborne systems. The smaller scale of these newer high resolution remote sensing systems allows scientists to characterize the geometric, spatial, and radiometric data properties using a single V&V site. The targets and techniques used to characterize data from these newer systems can differ significantly from the techniques used to characterize data from the earlier, coarser spatial resolution systems. Scientists are also using the SSC V&V site to characterize thermal infrared systems and active LIDAR systems. SSC employs geodetic targets, edge targets, radiometric tarps, and thermal calibration ponds to characterize remote sensing data products. This paper presents a proposed set of required measurements for visible through long-wave infrared remote sensing systems and a description of the Stennis characterization. Other topics discussed include: 1) The use of ancillary atmospheric and solar measurements taken at SSC that support various characterizations; 2) Additional sites used for radiometric, geometric, and spatial characterization in the continental United States; 3) The need for a standardized technique to be adopted by CEOS and other organizations.
The utility of a remote sensing system increases with its ability to retrieve surface temperature or radiance accurately. Research applications, such as sea temperature and power plant discharge, require a 0.2 C resolution or better for absolute temperature retrievals. Other applications, including agriculture water stress detection, require at least a 1 C resolution. To achieve these levels of accuracy routinely, scientists must perform laboratory and onboard calibration, as well as in-flight vicarious radiometric characterization. A common approach used for in-flight radiometric characterization incorporates a well-calibrated infrared radiometer that is mounted on a bouy and placed on a uniform water body. The radiometer monitors radiant temperature along with pressure, humidity, and temperature measurements of an associated column of atmosphere. On very still waters, however, a buoy can significantly distrub these measurements. Researchers at NASA's Stennis Space Center (SSC) have developed a novel approach of using an uncooled infrared camera mounted on a boom to quantify buoy effects. Another critical aspect of using buoy-mounted infrared radiometers is the need for extensive laboratory characterization of the instruments' radiometric sensitivity, field of view, and spectral response. Proper surface temperature retrieval also requires detailed knowledge of both the upward emission and the reflected sky emission. Recent work at SSC has demonstrated that the use of a polarization-based radiometer operating at the Brewster angle can greatly simplify temperature retrieval as well as improve overall accuracy.
In-flight measurement of spatial resolution were conducted as part of the ASA Scientific Data Purchase (SDP) Validation and Verification (V&V) process. Characterization included remote sensing systems with ground sample distance (GSD) of 1 meter or less, such as the panchromatic imager on-board the ICONOS satellite and the airborne ADAR System 5500 multispectral instrument. Final image products were used to evaluate the effect of both the image acquisition system (e.g., optics, electronics, motion, jitter, atmosphere) and image post-processing (e.g., resampling, modulation trasfer function (MTF) compensator). Spatial resolution was characterized by full width at half maximum (FWHM) of an edge response-derived line spread function. This was found to be a more robust measure of spatial resolution than the value of NTF at Nyquist frequency The edge responses were analysed using the tilted-edge technique that ovecomes the spatial sampling limitations of the digital imaging systems. As an enhancement to existing algorithms, the slope of the edge response and the orientation of the edge target were determined by a single computational process. Adjacent black and white square panels, either painted on a flat surface or deployed as traps, formed the ground-based edge targets used in the tests. Orientation of the deployable tarps was optimized beforehand, based on simulations of the imaging system. Numerous edge target images were analyzed for each of the tested sensors. The effect of such factors as acquisition geometry, temporal variability, MTF compensation, and GSD on spatial resolution were investigated.
The expected cost of adding thermal bands to the next generation Landsat Data Continuity Mission (LDCM) could be significant. This viewgraph presentation investigates both traditional cooled cross-track scanners and new architectures (cooled and uncooled) which could enable a low cost thermal capability. The presentation includes surveys of applications of Landsat thermal data and the architecture of thermal instruments. It also covers new thermal architecture sensor trades, ALI Architecture with Uncooled TIR Detectors, and simulated thermal imagery.
Methodologies for determining the modulation transfer function (MTF) and National Imagery Interpretability Rating Scale (NIIRS) of airborne panchromatic and multispectral digital imaging cameras using the NASA Stennis Space Center (SSC) edge target have been developed. The application of NIIRS for determining the relative utility of data sets against various applications is discussed. Sampling requirements and optimal flightlines for focal plane imagers have been investigated. It is shown that with sufficient flight tilt angle, an adequate number of samples can be obtained to define the edge response accurately. The relationships between the tilt angle, Ground Sample Distance (GSD), and the number of samples are discussed. Sample results using commercially available data and the NASA SSC edge target are presented and compared to theoretical results.