Appendix C - ETM+ and TM Cross-Calibration (Historical)
The following information is applicable to Landsat TM and ETM+ Pre-Collection data and is included for historical interest. Starting with Collection 1, Landsat 4-8 data are all radiometrically calibrated, regardless of sensor. (Micijevic, Haque, & Mishra, 2016) The historical information describes the Collection 1 cross-calibration that supersedes the old TM/ETM+ calibration.
The entire Landsat data record is important for terrestrial remote sensing and global change research because it covers a 40+ year period during which significant anthropogenic and natural terrestrial change has occurred. In order to maximize the benefits of this data record, steps are needed to ensure that the data are self-consistent and not significantly affected by artifacts of the various Landsat sensors. For the Landsat 7 mission, renewed efforts were made to ensure radiometric calibration across the whole Landsat series of sensors as well as with other Earth observation sensors such as Terra MODIS. A critical step in such a process is relating sensor radiometric calibration to an absolute scale, thus yielding image data at the top of the atmosphere (TOA) in physical units. Additional processing steps to retrieve surface parameters, such as reflectance and temperature, then become possible.
Consistency between the Landsat sensors starts with refined calibration of all individual sensors, including the development of a stable sensor (i.e., ETM+), detailed prelaunch characterization, as well as periodic on-orbit calibration. Post-launch radiometric calibrations are referenced to onboard standards and ground-based test sites with independent measurements. For Landsat 7 and ETM+, cross-calibration with earlier Landsat sensors begins by using near-simultaneous imaging of common Earth surface targets by Landsat 5's TM sensor. Typically, there is a limited overlap period when more than one of the sensors is operating. Such an overlap period with Landsat 5 was planned for the initial phases of the Landsat 7 mission. The resulting opportunity for radiometric cross-calibration between ETM+ and Landsat 5 TM during the early part of their many years of simultaneous operation is the main subject of this section. The following material was extracted and condensed from a paper that covers the subject in greater detail (Teillet, et. al., (2001).
C.2 Tandem Configuration
The launch of Landsat 7 on April 15, 1999 placed the spacecraft temporarily in an orbit very close to that of the Landsat 5 spacecraft. During this time, the mean altitude of Landsat 7 was 699 km, 6 km below the 705-km mean altitude of Landsat 5. At this altitude, the Landsat 7 ground track drifted slowly relative to the essentially fixed Landsat 5 pattern. The key period for the tandem configuration was June 1-4, 1999, when their tracks were almost exactly the same, but with a temporal offset on the order of only 10 to 30 minutes. This unusual and valuable opportunity was specifically designed to facilitate the establishment of data consistency comparisons between the ETM+ and TM sensors. During the tandem configuration period, image sequences corresponding to 791 matching scenes were recorded by both Landsat 7 ETM+ and Landsat 5 TM (Table C-1). Subsequently, the Landsat 7 orbit was adjusted for regular operations such that its 16-day repeat coverage cycle became offset from that of Landsat 5 by eight days. Given cloud cover and issues with data reception and recording, the number of useful data scene pairs acquired during the tandem phase was around 400 scenes.
The cross-calibration methodology documented in Teillet, et. al. (2001) is applicable to tandem image pairs acquired at other times and between other sensor pairs although it presents specific results for two different pairs of nearly coincident matching scenes from the Landsat 5 and Landsat 7 tandem configuration period. The main results were updated TM responsivities in the six solar reflective spectral bands referenced against the well-calibrated ETM+ responsivities in its corresponding spectral bands. The analysis approach included adjustments for differences in illumination as well as for differences in spectral response profiles between the two sensors.
C.3 Tandem Data sets Selected for Analysis
Attention was focused on two particular tandem image pairs for cross-calibration methodology development and analysis because of the availability of ground reference data. Both Landsat sensors imaged the Railroad Valley Playa (RVPN), Nevada on June 1, 1999, when a team from the University of Arizona made measurements of surface spectral reflectance and atmospheric aerosol optical depth the same day. Similarly, a team from South Dakota State University (SDSU) acquired the same types of ground reference data at a grassland test site near Niobrara, Nebraska (NIOB) on June 2, 1999, the day of the tandem Landsat overpasses for that site.
|Tandem Scene Coverage (June 1-4, 1999)|
|6/2 - 6/3 1999||95||65-87||ASA|
|6/3 - 6/4 1999||111||64-84||ASA|
Table C-1. Landsat 7 ETM+ and Landsat 5 Tandem Data Coverage
*Not all stations are still receiving, https://landsat.usgs.gov/igs-network
*Stations: ASA – ACRES, Alice Springs, Australia; COA – Cordoba, Argentina; CUB – INPE, Cuiaba, Brazil; FUI – ESA, Fucino, Italy; GNC – CCRS, Gatineau, Canada; JSA – Johannesburg, South Africa; KIS – ESA, Kiruna, Sweden; LBG - DLR, Libreville, Gabon; NOK – SI/EOSAT, Norman, Oklahoma; PAC - CCRS, Prince Albert, Canada; RSA – Saudi Arabia (for SI/Dubai)
Table C-2 provides information on the characteristics of the two data sets and Figure C-1 shows both Landsat image pairs. The RVPN test site is a dry-lake playa that is very homogeneous and consists of compacted clay-rich lacustrine deposits forming a relatively smooth surface compared to most land covers. The NIOB test site is characterized primarily by grasslands grazed by cattle and by a small area of agricultural crops.
|Railroad Valley Playa||Niobrara, Nebraska|
|Image Date||June 1, 1999||June 2,1999|
|Landsat 7 Offset from WRS||76.56 km East||18.15 km East|
|ETM+ Data Level||Level-0R||Level-0R|
|TM Data Level||Level-0||Level-0|
|ETM+ Solar Zenith Angle||24.28°||26.60°|
|TM Solar Zenith Angle||27.23°||28.67°|
|Terrain Elevation||1.425 km||0.760 km|
|Area Common to ETM+ and TM||10.7 km by 4.4 km||106 km by 66 km|
Table C-2. Characteristics of the Two Tandem Data Sets
*AOD550 represents aerosol optical depth at 550 nanometers
Figure C-1. Landsat 5 and Landsat 7 cross-calibration data sets of Railroad Valley Playa, Nevada (WRS-2 40/33) acquired June 1, 1999 and of Niobrara, Nebraska, (WRS-2 31/30) acquired June 2, 1999. Images are shown using Bands 5, 4 and 2.
C.4 Cross-Calibration Methodology
The cross-calibration methodology assumes that the Landsat 5 TM calibration is to be updated with respect to the Landsat 7 ETM+ sensor, which serves as a well- calibrated reference sensor (Barker et al., 1999). Because data acquisitions were only 10 to 30 minutes apart during the tandem configuration period, it is assumed that the surface and atmospheric conditions did not change significantly between the two image acquisitions.
Cross-calibration methodologies in general should consider adjustments as appropriate for Bi-directional Reflectance Factor (BRF) effects due to differences in illumination and observation angles. For the Landsat sensor image data pairs acquired during the tandem configuration period, the expectation is that such BRF adjustments are not necessary. The solar illumination geometries are very similar (within three degrees), satellite zenith angles are predominantly near-nadir, and relative azimuth angles between solar and satellite directions do not differ significantly from one Landsat overpass to the other. Nevertheless, it was necessary to address geometric, radiometric, and spectral differences between the sensors to account for their impacts.
Differences between the Landsat 7 and Landsat 5 sensors in their along-track and across-track pixel sampling made it difficult to establish sufficient geometric control to facilitate radiometric comparisons on a point-by-point and/or detector-by-detector basis. Therefore, the analysis approach made use of image statistics based on large areas in common between the image pairs. There are also significant differences in relative spectral response profiles between the corresponding ETM+ and TM spectral bands. The effects these spectral band differences have on measured TOA reflectances depend on spectral variations in the exo-atmospheric solar illumination, the atmospheric transmittance, and the surface reflectance.
C.5 Cross-Calibration Results
All of these challenges were overcome. As a result, detailed analyses were done on each corresponding band of data from these almost simultaneous image pairs acquired during the tandem configuration period. This allowed the use of the well-calibrated Landsat 7 ETM+ data to update the radiometric calibration of the Landsat 5 TM. The consistent results from the two field sites between their nearly coincident image pairs was encouraging. The band-to-band comparisons show negligible differences in spectral Band 4 to almost 4 percent difference in Band 7. The average difference is 1.6 percent, which, although based on only 12 spectral band cases, is a measure of the repeatability of the cross-calibration approach. Substantial differences between Landsat 5 TM's prelaunch calibration coefficients for the visible bands on-orbit were also detected through this analysis and updated coefficients were incorporated in subsequent processing of TM data.
The work indicates that the tandem cross-calibration approach provided a valuable "contemporary" calibration update for Landsat 5 TM's solar reflective bands relative to the excellent radiometric performance of Landsat 7 ETM+. Initial trials of the approach with two different tandem image pairs yielded repeatable results for TM responsivity coefficients that led to updated coefficients being used for processing of additional TM data sets. This effort, and other studies of sensor performance during the Landsat Project have been incorporated in the processing output products from the LPS.This type of work is ongoing and many significant results since this experiment can be found in the Appendix of Landsat publications or at the Landsat updates page.