SEARCH · Search NASA
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Descriptive and procedural background material for understanding results from radiometric calibration of the reflective bands and digital image products of the thermal mapper (TM) are discussed. The radiometric subsystems of the TM are described, with emphasis on the internal calibrator pulse shapes and timing cycle.
The processing concepts which form the basis of the Thematic Mapper (TM) Geometric Correction System are examined. The principle flight and ground segment subsystems are discussed. Correction data is generated and compared to TM image data. Geometric accuracy is defined and investigated along with a general system overview.
Thematic Mapper data processing during LANDSAT 4's first year was performed on an engineering evaluation basis. Fully corrected products were created for some 282 scenes during this period using software and systems based upon the intended full production systems to become operational following the evaluation period. The engineering systems included substantial software and procedures for assessing spacecraft, instrument and ground processing algorithmic behavior. The data systems are described in terms of performance objectives, processing organization and data flow, quality assurance measurements and achievement of goals. The a priori implementation of TM radiometric and geometric corrections is described. Changes to processing suggested by or implemented on the basis of on-orbit data analysis are discussed. Spacecraft, instrument and algorithmic performance are evaluated.
Data collected on the spectral characteristics of the LANDSAT-4 and LANDSAT-4 backup thematic mapper instruments, the protoflight (TM/PF) and flight (TM/F) models, respectively, are presented and analyzed. Tests were conducted on the instruments and their components to determine compliance with two sets of spectral specifications: band-by-band spectral coverage and channel-by-channel within-band spctral matching. Spectral coverage specifications were placed on: (1) band edges - points at 50% of peak response, (2) band edge slopes - steepness of rise and fall-off of response, (3) spectral flatness - evenness of response between edges, and (4) spurious system response - ratio of out-of-band response to in-band response. Compliance with the spectral coverage specifications was determined by analysis of spectral measurements on the individual components contributing to the overall spectral response: filters, detectors, and optical surfaces.
The results of a preliminary assessment of both raw and NASA processed Thematic Mapper (TM) data are discussed. Geometric correction of NASA processed TM data was carried out. Correction was possible to within 3 pixels in the along track direction and 2 pixels in the across track direction. A preliminary evaluation of TM imagery provided by the NASA LANDSAT Assessmemt System for geometric accuracy and map information content was performed on samples of imagery. The initial indications were that bands 3, 5 and 7 contain the most useful cartographic information. The resolution of rural and urban detail as well as the fit to plotted map detail was found to be improved over LANDSAT MSS, and such images may provide adequate revision information for 1:250,000 maps in areas where it is not currently profitable to use LANDSAT MSS. The relative gains and offsets for each detector in each band of raw data were calculated in a study of the radiometric correction of TM data. This was done for different subscenes as well as a full scene and the variation of the results with direction of scan and position of subscene were studied.
A technique for the radiometric correction of LANDSAT-4 Thematic Mapper data was proposed by the Canada Center for Remote Sensing. Subsequent detailed observations of raw image data, raw radiometric calibration data and background measurements extracted from the raw data stream on High Density Tape highlighted major shortcomings in the proposed method which if left uncorrected, can cause severe radiometric striping in the output product. Results are presented which correlate measurements of the DC background with variations in both image data background and calibration samples. The effect on both raw data and on data corrected using the earlier proposed technique is explained, and the correction required for these factors as a function of individual scan line number for each detector is described. It is shown how the revised technique can be incorporated into an operational environment.
Analysis was performed to characterize the radiometry of three Thematic Mapper (TM) digital products of a scene of Arkansas. The three digital products examined were the NASA raw (BT) product, the radiometrically corrected (AT) product and the radiometrically and geometrically corrected (PT) product. The frequency distribution of the digital data; the statistical correlation between the bands; and the variability between the detectors within a band were examined on a series of image subsets from the full scene. The results are presented from one 1024 x 1024 pixel subset of Realfoot Lake, Tennessee which displayed a representative range of ground conditions and cover types occurring within the full frame image. Bands 1, 2 and 5 of the sample area are presented. The subsets were extracted from the three digital data products to cover the same geographic area. This analysis provides the first step towards a full appraisal of the TM radiometry being performed as part of the ESA/CEC contribution to the NASA/LIDQA program.
The Thematic Mapper (TM) multispectral scanner system was placed into Earth orbit on July 16, 1982, as part of NASA's LANDSAT 4 payload. To determine temporal changes of the absolute radiometric calibration of the entire system in flight, spectroradiometric measurements of the ground and the atmosphere are made simultaneously with TM image acquisitions over the White Sands, New Mexico area. By entering the measured values into an atmospheric radiative transfer program, the radiance levels at the entrance pupil of the TM in four of the TM spectral bands are determined. These levels are compared to the output digital counts from the detectors that sampled the radiometrically measured ground area, thus providing an absolute radiometric calibration of the entire TM system utilizing those detectors. By reference to an adjacent, larger uniform area, the calibration is extended to all 16 detectors in each of the three bands.
In order to utilize the quantitative accuracy built into the Thematic Mapper effectively, more attention must be paid to calibration before launch, changes of calibration before launch, changes of calibration with time in orbit, and atmospheric interference with the measurements, especially in the 450 to 520 nanometer band. Recent experience with the Coastal Zone Color Scanner (CZCS) program has led to procedures wherein Rayleigh correction factors can be generated utilizing simultaneous surface truth data that empirically give correct upwelled surface radiances. Instruments, such as the CZCS, have shown that calibration changes first, and to the largest degree, at the shorter wave lengths, with lesser changes as wave length increases. These techniques are utilized to calculate a Rayleigh correction factor that, together with geometric terms, will give an accurate correction for this portion of the atmospheric contribution to the signal.
The performance of the Thematic Mapper (TM) as manifested by the quality of its image data is examined in order to suggest possible improvements for data production and assess the effects of data quality on its utility for land resources applications. The major emphasis is on the radiometric characteristics of TM data, with some attention to spatial and spectral characteristics.
A research program to measure the LANDSAT 4 Thematic Mapper (TM) modulation transfer function (MTF) is described. Measurement of a satellite sensor's MTF requires the use of a calibrated ground target, i.e., the spatial radiance distribution of the target must be known to a resolution at least four to five times greater than that of the system under test. A small reflective mirror or a dark light linear pattern such as line or edge, and relatively high resolution underflight imagery are used to calibrate the target. A technique that utilizes an analytical model for the scene spatial frequency power spectrum will be investigated as an alternative to calibration of the scene. The test sites and analysis techniques are also described.
Radiometric characteristics of the LANDSAT 4 Thematic Mapper (TM) that can be established without absolute calibration of spectral data are examined. The analysis is based on radiometrically raw (B type) data of three daytime and two nighttime scenes. A set of 512 lines are examined on an individual detector basis. Subscenes selected for uniform-radiance are used to characterize subtle radiometric differences and noise problems. Virtually all nonideal performance is incorporated into the fully processed (P type) images, but disguised by the geometric resampling technique.
LANDSAT Thematic Mapper P-data of Washington, D. C., Harrisburg, PA, and Salton Sea, CA are analyzed to determine magnitudes and causes of error in the geometric conformity of the data to known Earth surface geometry. Several tests of data geometry are performed. Intraband and interband correlation and registration are investigated, exclusive of map based ground truth. The magnitudes and statistical trends of pixel offsets between a single band's mirror scans (due to processing procedures) are computed, and the inter-band integrity of registration is analyzed. A line to line correlation analysis is included.
Geometric and radiometric characteristics of Thematic Mapper data are investigated through analysis of linear features in the data. A linear feature is defined as two close, parallel and opposite edges. Examples in remotely sensed data are such features as rivers and roads. The geometric and radiometric precision TM data is sufficient to allow accurate measurement of linear feature widths. Results also confirm a 28.5m ground IFOV as specified prior to launch. The increase dimensionality of the TM data as compared with MSS data allows the possibility of independent verification of results by using data from several bands.
The potential effects of high frequency vibrations on the final Thematic Mapper (TM) image are evaluated for 26 scenes. The angular displacements of the TM detectors from their nominal pointing directions as measured by the TM Angular Displacement Sensor (ADS) and the spacecraft Dry Rotor Inertial Reference Unit (DRIRU) give data on the along scan and cross scan high frequency vibrations present in each scan of a scene. These measurements are to find the maximum overlap and underlap between successive scans, and to analyze the spectrum of the high frequency vibrations acting on the detectors. The Fourier spectrum of the along scan and cross scan vibrations for each scene also evaluated. The spectra of the scenes examined indicate that the high frequency vibrations arise primarily from the motion of the TM and MSS mirrors, and that their amplitudes are well within expected ranges.
The design of the Thematic Mapper (TM) multispectral radiometer makes it susceptible to band-to-band misregistration. To estimate band-to-band misregistration a block correlation method is employed. This method is chosen over other possible techniques (band differencing and flickering) because quantitative results are produced. The method correlates rectangular blocks of pixels from one band against blocks centered on identical pixels from a second band. The block pairs are shifted in pixel increments both vertically and horizontally with respect to each other and the correlation coefficient for each shift position is computed. The displacement corresponding to the maximum correlation is taken as the best estimate of registration error for each block pair. Subpixel shifts are estimated by a bi-quadratic interpolation of the correlation values surrounding the maximum correlation. To obtain statistical summaries for each band combination post processing of the block correlation results performed. The method results in estimates of registration error that are consistent with expectations.
An initial analysis of LANDSAT 4 Thematic Mapper (TM) data for the discrimination of agricultural, forested wetland, and urban land covers is conducted using a scene of data collected over Arkansas and Tennessee. A classification of agricultural lands derived from multitemporal LANDSAT Multispectral Scanner (MSS) data is compared with a classification of TM data for the same area. Results from this comparative analysis show that the multitemporal MSS classification produced an overall accuracy of 80.91% while the TM classification yields an overall classification accuracy of 97.06% correct.
Thematic Mapper (TM) data from Mississippi County, Arkansas, and Webster County, Iowa, were examined for the purpose of evaluating the image data quality of the TM which was launched on board the LANDSAT-4 spacecraft. Preliminary clustering and principal component analysis indicates that the middle infrared and thermal infrared data of TM appear to add significant information over that of the near IR and visible bands of the multispectral scanner data. Moreover, the higher spatial resolution of TM appears to provide better definition of the edges and the within variability of agricultural fields. The geometric performance of TM data, without ground control correction, was found to exceed expectations. The modulation transfer function for the 1.65 m band was found to agree with prelaunch specifications when the effects of the GSFC cubic convolution and the atmosphere were removed. The band to band registration for the bands within the noncooled focal plane was found to be better than specified. However, the middle infrared and thermal infrared, which are on a separate cooled focal plane were found to be misregistered and were significantly worse than prelaunch specifications.