History of Remote Sensing: Landsat (ne: Earth Resources Technology Satellite)

---

The sources of many of the images displayed throughout this Tutorial are ERTS-1 (renamed Landsat) and subsequent Landsats, earth-observing satellites originally developed by NASA and operated by them and/or several other agencies from 1972 through the present (Landsat-7). The Landsat program is described and the orbital and instrumental parameters for this series of satellites are given. Discussion of Landsat's two prime sensors, the MSS and TM, and examples of images produced from each is deferred to the next 6 pages.

---

History of Remote Sensing: Landsat (ne: Earth Resources Technology Satellite)

By the late 1960s, the first unmanned satellite specifically dedicated to multispectral remote sensing entered the planning stages. NASA carefully designed and constructed, then launched ERTS-1 (Earth Resources Technology Satellite) on July 23, 1972.

No satellite program designed to monitor the Earth's surface, the features thereon, and thus the environmental states that relate to land cover, land use, and natural resources, has proved more important than ERTS (Earth Resources Technology Satellite). The key people involved in its conception, inception, and operation deserve special mention here:

William T. Pecora was Under Secretary of the Interior and Landsat program champion. Unfortunately, he died just a few days before Landsat-1 was launched.

William A. Fischer, one of the nation�s first photogeologists, had the original idea for Landsat, which he shared with Pecora.

Dr. Arch Park chaired the management team that defined the U.S. Department of Agriculture�s research program for Landsat.

Dr. William Nordberg designed NASA-supported research programs for Landsat, coordinating the activities of 300 scientists from 38 countries during Landsat�s first year.

John De Noyer directed the development of NASA-supported research programs for Landsat.

Bruton �Doc� Schardt, Landsat�s first program director, worked tirelessly to ensure the simultaneous development of hardware and research programs, while resolving numerous political and budgetary issues.

Virginia Norwood, senior scientist at Hughes Aircraft Company, headed the team that developed the Landsat I MultiSpectral Scanner

All were friends or acquaintances of the writer (NMS) and William Nordberg was responsible for bringing me into remote sensing and the Earth Observing program.

An interesting, "off the cuff", synopsis by Stephen Hall of the genesis and early days of the Landsat program can be visited at a USGS website called EarthShots.

Renamed Landsat, ERTS-1 was the first in this series (seven to date) of Earth-observing satellites that have permitted continuous coverage of most of Earth's surface since 1972. Launch dates are: Landsat 1, July 23, 1972; 2, January 22, 1975, 3, March 5, 1978; 4 , July 16, 1982, 5, March 1, 1984 (Landsat 6, launched later, failed to operate); and 7, April 15, 1999. (As of May 1999, only Landsats 5 and 7 are still operational, i.e., acquiring data; the older ones have been shut down.). To help to visualize this information, we repeat an illustration first shown on page 1 of the Overview:

The orbital conditions for Landsat 1 are depicted in this illustration and described in the next paragraph.

The first three Landsats orbited at an altitude of 570 miles (923 km); 4, 5 and 7 at 435 miles (705 km). The orbits of all Landsats are near-polar (inclined 9.09° from a longitudinal line) and Sun-synchronous (pass every time over the equator between 9:30 and 10:00 AM), making 14 passes in descending mode (southward from the North pole in the daylight mode) each day (about 103 minutes for a complete orbital circuit). After any given orbit, the spacecraft will occupy its next orbit some 1775 miles (2875 km) to the west; on the next day, the orbits are configured so that orbit 15 has displaced westward by 98 miles (159 km) at the equator. Landsats 1-3 will reoccupy almost precisely the same orbit after 252 such orbits, or 18 days later; Landsats 4, 5 and 7 reoccupy on a 16 day cycle. Under the above orbital conditions, and with an angular field of view if 11.58 ° the width of a Landsat MSS scene is 185 km (114 statute or 100 nautical miles). The continuous imagery along an orbital strip is cut (subdivided) every 185 km to produce a given image length. (Thus an image is equi-dimensional at 185 x 185 km.) These same frame dimensions hold for the Landsat Thematic Mapper (TM) images, discussed later.

The orbits of the Landsats are termed "paths" and the location of individual images along these paths are fixed by a row system. This Worldwide Reference System is described in some detail at this U.S. Geological Survey web site.

I-19 Approximately how many square miles are enclosed are enclosed in a Landsat MSS frame? Square kilometers? Acres? (Note: there are about 640 acres in a square mile.) ANSWER

All the Landsats follow a near-polar orbit (inclined about 98° to the equator; passing within 8° of the poles) and are sun-synchronous, meaning the orbit precesses about Earth at the same angular rate as Earth revolves about the Sun. Thus, it crosses the equator (traveling from North to South) each day between 9:30 and 10:00 A.M., local time. Landsats 1-3 make 14 full orbits (each successive one displaced 2875 km [ 1785 miles] to the west) each day (three over the U.S.), and after 252 orbits in 18 days, they repeat their previous ground tracks. Landsats 4, 5, and 7, from a lower altitude (705 km [438 miles]), cover the same ground track again every 16 days, after 233 orbits. It takes about 11,000 scenes to fully image the entire Earth's land surface (except for polar regions).

Since the first Landsats, their success and the obvious value of the information they send back have prompted many countries to set up direct-readout receiving stations so as to obtain the raw data in near real time without depending on the NASA processing centers. These stations are charted by NASA and must agree to distribute the data to users in their region. Here is a map of currently active stations:

In the U.S., the primary receiving station for Landsats 1-3 was at Goddard Space Flight Center. Later Landsats were operated by different organizations and another major receiving station was established at the United States Geological Survey's EROS Data Center in Sioux Falls, S.D. Processing at these places include reformatting of the raw data that involves orbital, geometric, and radiometric corrections. For years, the output data were distributed on 9-track Computer Compatible Tapes (CCTs) but today data for Landsat and the EOS satellites are put on 8 mm, 4 mm, DLT, CD-ROM and 3 1/2 disk storage units.

The first three Landsats used two sensor systems: the Return Beam Vidicon (RBV) and the Multi-Spectral Scanner (MSS). The RBV consisted of three TV-like cameras which used color filters to provide multispectral images in EM bands centered in the blue-green, yellow-red, and red-IR. This sensor failed early on the ERTS-1 and never came into routine use, although it flew again on Landsat-2. Below is an example of the red-IR band covering northern New Jersey and including the New York City region.

Landsat-3 carried a four-camera RBV array, with each being a panchromatic (0.505 - 0.750 µm) imager, that provided four contiguous images at 30 m (98ft) resolution; each image comprises approximately one-quarter of a full Landsat MSS scene). This Landsat-3 RBV scene shows Cape Canaveral, Florida where the Space Shuttle launches.

Landsats 4, 5 (and 6) carried, in addition to the MSS, a new instrument called the Thematic Mapper. Here are an artist's sketch of the Landsat 4 spacecraft and a drawing with labeled components, which point to the notable differences from the previous Landsats:

Landsat 7 mounts only a single payload, the Enhanced Thematic Mapper (ETM+). The ETM+ includes not only the 7 TM bands (the thermal band [6] has 60 m resolution; all others achieve 30 meter ground resolution), but also a panchromatic band that is capable of 15 m resolution. This artist's rendition shows that the spacecraft differs in general construction from the previous "birds" (a term of familiarity used by the launch team).

Compare this view of Landsat 7 in its assembly room with the view of Landsat 1 at the top of the page:

As of this writing Landsats-5 and 7 are still operational and returning data.

These and other earth-observing satellites send copious amounts of data to various receiving stations each day (one estimate puts it this way: The data returned over one complete 16 day repeat cycle would be enough to fill 21 Encyclopedia Brittanicas). Processing of raw data into formats preferred by the user community usually is done at the station site. The original receiving station and processing facility was at NASA's Goddard Space Flight Center, outside Greenbelt, MD. The prime facility for current data is now the LP DAAC (Land Processes Distributed Active Archive Center), a part of the U.S. Geological Survey's EROS Data Center, east of Sioux Falls, South Dakota. Its Internet home page is worth a visit. Older Landsat data are there in part but also at other facilities.

The subject of primary data processing, archiving and distributing of Landsat material is extensive and diverse. The writer (NMS) deems it beyond the scope of this Tutorial. But, just to hint at the complexity, glance at this illustration - a flow chart for Landsat-7 data handling at the EROS DAAC:

The EOSDIS system now includes data distribution from Terra and Aqua, and other satellites.

For the intrepid (with time to burn), a visit to the Landsat program historysite developed at NASA's Ames Research Center will enlighten as to the checkered but distinguished history of this premier group of satellites.

---