| Landsat Technical
Characteristics
ES 351, ES 771, ES 775
James S. Aber |
Landsats 1-3 carried return-beam vidicon (RBV) cameras and multispectral scanners (MSS). Landsats 4-5 carried the MSS as well as the thematic mapper (TM), a more advanced scanner. Landsat 7 has an enhanced thematic mapper plus (ETM+).
Landsat orbit configuration
Landsat satellites follow near-polar orbits that are inclined approximately 8° to
the poles. Landsats 1, 2 and 3 orbited at 920 km altitude 14 times per
day. Over a period of 252 orbits, or every 18 days, a satellite covered
every portion of the Earth, except for polar regions above 82°
latitude. Landsats 4, 5 and 7 orbit at 705 km altitude with a 16-day repeat
interval for global coverage. The Landsat orbit is designed to be sun
synchronous. In other words, the satellite crosses the equator at the
same local sun time with each orbit. This provides for some uniformity of
lighting conditions.
 |
| Landsat orbit pattern and path numbering scheme for Landsats 4, 5 and 7. Note the variation in path orientation according to latitude. The curved paths result from the map projection. Taken from U.S. Geological Survey Landsat 7 homepage. |
Scanner capabilities
The multispectral scanner (MSS) detects reflected solar energy in two bands
of visible light and two bands of short-infrared energy. Scanning is
accomplished by a mirror that oscillates back and forth across the scene.
Energy is measured in four bands for six lines with each scan of the
mirror. Energy intensities are sampled every 57 m along scan lines, and
the data are recorded as digital values: 0-127 for bands 1, 2 and 3; 0-63
for band 4 (Holkenbrink 1978). The thematic mapper operates in seven bands
of visible and infrared energy with a resolution of 30 m. Data are
recorded on byte-binary scale (0-255). The ETM+ adds a panchromatic band
(0.5-0.9 µm) at 15 m resolution.
A complete image takes about 25 seconds to scan, during which time the
Earth rotates eastward beneath the satellite path. To compensate for this,
scan lines are slightly offset in the resulting image, which is shaped like
a parallelogram. The orientation of a scene depends on latitude; at about
50° the scene is tilted approximately 10° from true north.
Characteristics of Landsat MSS digital data
The characteristics of Landsat MSS data have changed over the years. Archive Landsat MSS datasets are supplied from the EROS Data Center in two formats.
- X-format, 1972 to 1979. Band-interleaved by pixel pair (BIP-2).
Rectangular pixel shape and paired interleaving format proved awkward, so this
format was discontinued in 1979. Data quality varies substantially, and many
scenes have incomplete data. X-format is recommended only for environmental
change applications in which data from the 1970s are necessary.
- EDIPS (EROS Digital Image Processing System) format is fully processed data. The raw data are resampled so the each cell is 57 x 57 m square. A geometrically corrected, fully processed Landsat MSS image nominally contains 2983 scan lines with 3596 pixels per line. The data are available in band-sequential (BSQ) or band-interleaved-by-line (BIL) format. In BSQ format, which is most convenient for image processing, each band is about 10 MB in size. EDIPS format is standard for all MSS data beginning in 1979; data quality is fair to excellent.
Landsat MSS datasets are now considered largely historical in nature, as the TM and ETM+ instruments provide greater spatial and spectral resolution. However, MSS datasets are often necessary for temporal change analysis going back to the 1970s--see
From the metadata listing, digital datasets are chosen to order. Archive Landsat digital data (and film images) are held at EDC and at several international remote sensing centers. The current cost of digital datasets ranges from $200 (raw MSS) to $600 (georegistered ETM+) per scene on CD or DVD from the EDC, U.S. Geological Survey.
Careful selection of scenes is most important for all types of geological,
hydrological, geomorphological, and other environmental applications. Only
images with the highest quality rating and minimal (<10%) cloud cover
should be considered. Time of year is an important consideration, as it
affects vegetation, soil moisture, agriculture, sun elevation, snow/ice
cover, and other seasonal conditions.
Return to Landsat remote sensing.