The main camera on Landsat 1 functioned for about two weeks.
The satellite, which was launched on July 23, 1972, carried two instruments. One of these was a camera of a television type, which all people anticipated would be the one to carry out the main task. The other was an experimental scanner, which many researchers did not trust. The camera was finally switched off by early August, after which the instrument that had been doubted became the core of the mission.
In the previous GIS Beats article about Sputnik 1, we examined a satellite which had not only no camera but also no sensors of any kind. The story continues with Landsat 1, which was the first satellite to be launched with the particular aim of studying the Earth's land and which almost suffered a failure during its journey. So what does the 1972 near-miss have to do with you today? Well, the decisions taken when launching that spacecraft still form the basis of the remote sensing work that we carry out every week, and the figures also offer a helpful lesson on scale for anyone using drones.
What was wrong with the Landsat 1 camera?
The instrument everyone trusted
The Return Beam Vidicon, or RBV, was intended to be the highlight of the project, having been developed by RCA and being, in effect, a television camera derived from the tube technology used in 1960s weather satellites, with three cameras taking pictures at the same time in three different spectral bands. It performed well at first: according to NASA reports, it provided about 1,400 high-resolution images in its first two weeks.
On August 6, a major power surge occurred. The spacecraft lost its ability to control its orientation and turned away from Earth, whereupon the ground controllers were forced to switch off the RBV. Later ground tests found that the cause of the problem had been a defective inductor in the power-switching circuit. When safety concerns together with the limited amount of tape storage were taken into account, the RBV was finally permanently switched off just 15 days after launch.
The instrument nobody trusted
The backup system in question was the Multispectral Scanner System (MSS), and it was an experimental one. Rather than capturing a full-frame image, it employed a moving mirror to sweep across the ground and to record digital data. At that time, many researchers were doubtful about the system's ability to achieve the same quality as the camera systems employed in aerial studies, as NASA's page on Landsat spectral bands explains. The system had already been tested on Earth but had never been used in the severe conditions of space.
The entire mission then came down to the instrument about which people were least certain. Both NASA and the USGS state exactly what happened afterwards: the MSS data proved to be clearly superior to that of the camera.
How did the multispectral scanner on Landsat 1 function?
Allow me to explain this in simple terms. A camera takes in the entire frame at one time, whereas a scanner assembles the image in strips. With regard to the MSS, a small beryllium mirror was moved back and forth between two positions, reflecting light from the ground to the detectors. Each pass contributed six new lines to the image while the satellite moved forward along its path, as stated in NASA's history write-up. The entire instrument was small, measuring about 89 cm in length, 59 cm in width, and 40 cm in height, and weighed approximately 48 kg, according to NASA's profile of its designer.
Virginia Norwood, who was an engineer at Hughes Aircraft, is referred to by NASA as the 'Mother of Landsat'. Although she had designed a six-band scanner, the one that was flown on Landsat 1 contained only four bands, as her biography states.
What made those four bands important?
The green, red, and two near-infrared bands were used. NASA refers to the MSS as having a coarse spectral resolution, essentially being a digital equivalent of the colour infrared film which had been commonly used since World War II. Now for the interesting point: healthy plants absorb red light and reflect a great deal of near-infrared. When you compare these two bands, vegetation becomes clearly distinguishable from all the surrounding areas. This contrast is the basis of NDVI, and we have already discussed in our post why NDVI alone doesn't tell you if a crop is actually stressed where it does and where it fails. The bands on the 1972 scanner were already focusing on the same concept.
What actual achievements did Landsat 1 have?
The satellite was intended to have an operational life of one year, and it continued to function until 6 January 1978, by which time it had imaged approximately 75% of the Earth's surface, as stated by USGS. It orbited the Earth in a polar, sun-synchronous trajectory at an altitude of 917 km, repeated its coverage every 18 days, and had a swath width of 185 km. During its mission, the MSS acquired more than 175,000 scenes at a resolution of 80 m, according to NASA's mission page.
Within a few days of its launch, it also detected a fire of 81,000 acres burning in central Alaska, an area of about 328 square kilometres.
The journey was far from smooth since, in 1972 and 1974, failures of the tape recorders caused the satellite to have to depend on real-time downlinks to ground stations for the remainder of its operational life, and it continued to send the data anyway.
A brief comment regarding the numbers: although the USGS gives the resolution of the first Dallas–Fort Worth image as 60 metres per pixel, NASA states that the MSS has a resolution of 80 metres. In this text we adopt NASA's nominal value of 80 metres. It should be understood that a sensor's nominal resolution and the pixel size of a processed product are related but distinct figures.
How Did Free Landsat Data Change Remote Sensing?
Making Landsat data free in 2008 changed remote sensing in big ways. Before that, high costs limited who could use these satellite images. Only governments, large companies, or researchers with big budgets could afford them. Once the data became free, more people from different backgrounds started using it. Scientists, students, nonprofits, and small businesses could all explore satellite images for their own work.
This shift led to more research and new ways to use remote sensing. People used Landsat images to watch forests, measure crops, and track urban growth. They also used the data to study water resources and climate change. Many new tools and online platforms appeared to help users work with satellite images. These tools made remote sensing easier to use and understand.
Free Landsat data also helped with global projects. Groups could now map land use, spot wildfires, and respond to disasters faster. Open access to the data made it easier for people in different countries to work together. As a result, free Landsat data helped more people use remote sensing to solve real-world problems.
Half the story was the construction of the satellite and the other half was getting the data to people. For many years, Landsat scenes carried a fee. In the fiscal year 2001, which was the busiest year before the policy changed, the USGS provided 25,000 scenes. The USGS announced its policy of making the data free and open on April 21, 2008, and by the fiscal year 2009 more than one million scenes had been downloaded, as stated by USGS EROS.
Here’s another way of looking at it: when there was a price tag, about 53 scenes a day left the archive, but once the price tag had been removed, the number rose to 5,775 a day, as the USGS reported ten years later.
What does that actually mean? In the old days, you would buy a few scenes. Now you can view every scene taken at a single location over many decades and see how the land has changed. As one former chief scientist at EROS put it, people were now using "the data they needed, not the data they could afford". That represents a major shift in the way the field operates.
If you would like to find out where you can obtain this type of imagery nowadays, we have compiled a practical guide to free satellite imagery sources for 2026.
Why does a pixel size of 80 m still matter to drone surveyors?
Since we mostly use drones, I find it useful to compare the Landsat 1 figures with our own. During a survey of a tail race corridor for hydropower infrastructure in northern Pakistan, we flew at an altitude of 130 metres above ground level, took 837 images over an area of 0.561 km², and processed them to a ground sampling distance of 3.95 cm per pixel.
The nominal resolution of Landsat 1 was 80 metres. That works out to be about 2,000 times coarser. To put it simply, the entire corridor would be contained within around 88 MSS pixels. (Although GSD and a satellite sensor's nominal resolution are related but not identical, the statement should be understood as giving a sense of scale, not as a direct comparison.)
If you look at the other side of the trade, Landsat 1 covered a swath 185 km wide from an altitude of 917 km, that is about 7,000 times higher than our flight altitude, and it repeated this coverage every 18 days.
What you should do then is begin by asking the question. In the case where you want to find out if there has been a change in vegetation over a ten-year period, a satellite archive is the appropriate tool; but if you need contours, volumes, or an orthomosaic for a particular site, a pixel of that size won't suffice, which is precisely where drone imaging and processing comes into play.
What Landsat 1 Still Teaches Us
I see two lessons from this situation. The first is that the instrument that had been doubted ended up being the mission, and that serves as a useful reminder to always keep an experimental option available. The second point is that the sensor alone was not the whole value; it was access to the data that transformed Landsat from a satellite into a full area of research.
The programme has continued to operate since then, with Landsat 9 being launched in 2021 via this EarthDate episode. The next time you open a scene and notice healthy vegetation appearing bright in the near-infrared, think back to its origins: a camera which failed after approximately two weeks and a backup whose reliability no one was certain about.
If you're beginning your study of remote sensing, then my advice is straightforward: master one sensor, make sure you understand its pixel size, and select your data according to the question you want to answer.
