Part I: The road so far
Commercial Earth observation is undergoing rapid changes: much like rocket companies such as SpaceX, Blue Origin and a handful of others are challenging the established players in the launch industry, new ways of doing things are being introduced by “new space” companies in the commercial Earth Observation business. Can they compete with, and eventually replace the traditional satellite builders and operators? Or will they address different markets and coexist? In this series of posts, I will give my analysis on where I think this industry is going, starting with this first entry on the history of commercial optical earth observation.
To understand better what is changing in satellite earth observation, let’s first look back to where it came from.
Optical earth observation satellites have existed since 1959, just two years after the first artificial satellite reached orbit. That year, the USA launched Vanguard 2, the first weather satellite. It carried an optical scanner to study cloud cover, and sent its images back to Earth through radio link. It started a long line of American weather satellites, leading to today’s NOAA satellites. However, the images collected by these systems do not provide much detail: even today, they have resolutions of a few hundred meters. This is fine for meteorology but not for other uses.
The same year Vanguard 2 was launched, the USA also launched the first spy satellites. Several generation were developed, starting with the Corona program. These satellites used a film system to capture the images. The films were then stored in a recovery capsule, after which the capsule was deorbited and recovered. The advantages of using film together with large optics were the high resolution of the images, which was measured in feet, and the removal of the storage and transmission electronics. This design choice however resulted in a relatively long time between the acquisition of the image and its delivery, and required a high launch cadence and complex recovery operations. The USSR had a similar program of film-based reconnaissance satellites.
The imagery produced by the spy satellites remained classified, and so did most of their technology. In the USA, NASA even had an agreement with the Department of Defense not to develop sensors with better than 20 meter resolution. Nevertheless, civilian medium-resolution satellites were eventually developed, using electronic sensors and radio downlink of images.
At first, these systems, such as the first satellites of the American Landsat program, could only provide images with resolutions in the tens of meters. They were best suited to mapping and agricultural uses. The Spot 1 satellite, launched in 1986 by the French, was the first satellite providing commercial imagery at 10m resolution. Two month after its launch, it took a picture showing the damaged nuclear reactor at Chernobyl, confirming the explosion and the usefulness of timely, higher-resolution commercial imagery.
The Landsat and Spot programs were still cutting-edge technology at the time, leading to high costs. However, the developments in electronics made them affordable for the civilian space programs of industrialized nations. The satellites themselves were heavy -around 2t-, and had limited resolution, especially when compared with airborne imagery. They also had limited agility, meaning they could point in a direction to take an image, but not repoint quickly in another direction to take another image. Still, they found a market, leading to the launch of improved commercial satellites, with a new design, in the 2000s.
Current state of the industry
Encouraged by the success of Spot, and by the fact the Russian started selling high-resolution images from their military film-based systems in the early 90s, many countries developed even higher-resolution civilian satellites. The USA passed the Land Remote Sensing Policy Act in 1992, which allowed private US companies to operate earth observation satellites. This led to the creation of DigitalGlobe and GeoEye. Starting from the late 90s, the two companies launched several high-resolution, high-agility satellites, eventually offering 50 cm imagery on the commercial market, and even higher-resolution to the Pentagon. This regulatory limit on resolution was progressively relaxed over the years, to keep the US industry competitive with foreign companies.
Several countries followed suit. As the satellite pictures at the top of this paragraph show, they adopted similar solutions to solve the same problem: all those satellites have relatively large telescope apertures to provide high resolution images, but are very compact to be able to quickly repoint themselves to take another image. The pointing agility is a critical feature of these designs, because high-resolution telescopes have a low field of view: they can only produce images which are around 15 km wide. So to image two places located in the same region but more than 15km apart, they have to repoint and take a new image. The faster they can repoint, the more interesting places they can image, and the more value they generate.
The technology required to build these satellite still made them expensive, coming at a few hundred millions dollars a piece, but made them more accessible than the previous generation. Several countries, such as South Korea, Israel, India and the UK, joined the club of high-resolution satellite manufacturers. A dynamic export market emerged and is still growing, as countries and companies without the technology to build such a satellite bought one from a foreign manufacturer.
Thanks to the advances in electronics miniaturization, the barrier to entry in the satellite design and manufacturing business has been greatly reduced. New entrants, such as Terra Bella (formerly Skybox Imaging), Planet Labs, and BlackSky Global have launched observation satellites, or announced they will do so. In order to better understand what kind of market they aim for, the next post will focus on the different uses of satellite imagery.
2 thoughts on “The Future of Optical Earth Observation”
[…] observation satellites do not come cheap. As mentioned in the previous post of this series, until a few years ago, they could cost a few hundred million dollars a piece to […]
[…] post is a part of a series on the future of earth observation, and is a follow up to The Future of Optical Earth Observation I: The road so far and The Future of Earth Observation II: What do users really want, […]