There are essentially three types of Earth orbits: high, medium and low Earth orbit. Satellites that orbit in a medium (mid) Earth orbit include navigation and specialty satellites, designed to monitor a particular region. Most scientific satellites, including NASA’s Earth Observing System fleet, have a low Earth orbit. On which orbit a satellite will be launched to, depends mainly on its application. The orbit types can be categorized according to their height.
The orbit height of a satellite corresponds to the distance between the Earth’s surface and the satellite. It determines its speed as it rotates around the Earth. Due to Earth’s gravity, the pull of gravity is stronger for lower orbits than for higher orbits. Therefore, a satellite situated on a lower orbit will circle the Earth faster than a satellite situated on a higher orbit.
High Earth orbit: it describes orbits situated at about 36000 km above the Earth’s surface (42164 km from the Earth’s center). At this exact distance, the speed of the satellite on the orbit matches the Earth’s rotation, i.e. the satellite needs 24 hours to complete a full rotation on the orbit, when the orbit is situated exactly above the equator. Such orbits are also called geosynchronous orbits, as the satellite moves at the same speed than the Earth and seems to stay in place over a specific location. Those orbits are mainly used for weather and communication satellites
Medium Earth orbit: it describes orbits situated at about 20200 km of the Earth’s surface, or 26560 km of the Earth’s center. At this height, a satellite rotates twice around the orbit during one Earth’s rotation. This orbit is also called semi-synchronous and this is the orbit type used by Global Navigation Satellite Systems such as GPS and GLONASS. A further important medium Earth orbit is the Molniya orbit which allows the observation of the poles, otherwise nearly impossible with equatorial geosynchronous orbits.
Low Earth orbit: this type of orbits are used from almost all dedicated scientific Earth Observation satellites. Most of them use a particular, nearly polar orbit inclination, meaning that the satellite rotates around the Earth nearly from pole to pole (instead of around the equator as it is the case for geosynchronous satellites). This rotation takes about 99 minutes, depending of the specific orbit inclination. During one half of the orbit, the satellite views the daytime side of the Earth, i.e. the illuminated side. At the pole, satellite crosses over and views the nighttime side of Earth. Back to the daylight side, the satellite can view the area adjacent to the region flown over in the last orbit path, due to the simultaneous Earth’s rotation. In 24 hours, satellites situated on these orbits view almost all the Earth twice, for optical satellites once in daylight and once in the dark. Radar satellites seen each Earth region twice, from two different illumination directions. These specific polar-orbits are called sun-synchronous, as the local solar time stays the same each time a satellite flies over a specific region. This has the advantage of providing an almost constant angle of sunlight for each region on the Earth’s surface viewed by the satellite over time and ensure repeatable sun illumination conditions; the angle will only vary seasonally due to the Earth revolution around the sun. Due to this consistency, images of a specific region would not show much illumination changes due to shadows or sunlight and image interpretation over time such as change detection or monitoring approaches are possible. Because a sun-synchronous orbit does not pass directly over the poles, there is a data gap over both poles where no data is acquired.
The choice of a satellite orbit mostly depends on its main application. From this point of view it represents a crucial part of a satellite mission design. The most important parameters to describe a satellite orbit are the inclination angle i (of the orbit plane respect to the equatorial plane) its eccentricity e and its height H from the Earth's surface. In principle whatever eigth H can be used, provided that the speed of the satellite on its orbit allows the centrigugal force to exactely compensate the gravitational one at that heigth. Polar (i close to 90°) and Geostationary (i=0, H=35.800 km) orbits are the most common choices for EO satellites. In principle one single polar satellite can be sufficient to guarantee the global coverage of the Earth with equal quality of the images at all latitudes. All Geostationary satellites share the same circular orbit with H around 36000 km where the required speed exactely correspond to the one required to travel an entire orbit in 1 sideral day (orbital period P = 1 sideral day). This means that the satellite footprint is permanently in place over a specific Earth's location (e.g. for Meteosat 0°N, 0°E) allowing a quasi-continuous monitoring of a whole Earth's emisphere (with poor visibility of Earth's edges including Poles). Polar satellites' heigths are usually in between 700-800 km, with orbital periods around 100min (i.e. about 14,5 orbits/day) even if, lower orbits are also chosen particularly for very high spatial resolution payloads. Lower inclinations are also used (quasi-polar orbits) for specific applications. Due to the asphericity (and mass inhomogeneity) of the Earth, satellite orbit plane rotates around the Earth's polar axis with a period Pp producing (for elliptical orbits) the rotation of the orbit itself in its plane. A common choice for most EO polar satellites is to choose the orbital parameters in a way that Pp=1 year (Sun-Synchronous orbits). Due to the synchronism between Earth's revolution around the Sun and the orbit plane precession around Earth' axis, satellite passages happens at the same local solar time (similar illumination conditions) each time it flies over a specific region. This ensure repeatable sun illumination conditions facilitating image interpretation particularly for change detection or land monitoring applications. Other choices are possible when it is required to monitor with continuity high latitude regions.
This is the case of Molniya orbits which combine the continuity of observations typical of geostationary satellites with the possibility, offered by polar orbits, to overfly the highest latitudes regions. Its characteristics are: high eccentricity (e.g. e=0,74, axes 500 and 23.000 km), P=1/2 sideral day (Geo-Synchronous), inclination (i=63,4° or i=116,6°) which guarantees the satellite footprint at the apogee remaining positioned on a fixed ground point (non-rotating orbit). This way the satellite will spend more than 93% of its orbital period looking to the same emisphere even from a high latitude point of view.
So called altimetric orbits respond to the specific needs of altimetry. In this case the orbital parameters are chosen in order to guarantee, for example: a) that the ascending and descending sub-satellite tracks intersect at roughly 90 degrees on the Earth’s surface (so that orthogonal components of the surface slope can be determined with equal accuracy; b) the possibility to monitor all phases of tidal effects on ocean surface.
Particularly important for several applications (multi-temporal analyses, change detection, etc.) are the Exactly repeating orbits.
They are conceived in order that the sub-satellite track will repeat itself exactly after a certain interval of time. This allows images having the same viewing geometry during the satellite’s lifetime making moreover available a particularly simple method of referring to the location of images (navigation or geo-referenciation) for example by referring to a ‘path and row’ system used for instance by the Landsat World Reference System (WRS). It is possible to arrange satellite orbits parameters in order to contemporary guarantee the sun-syncronism so that, not only satellite images collected on the same region can be easily super-imposed each-other but the same illumination and viewing geometry can be achieved. This is, for instance, the choice adopted for LANDSAT satellites whose images are typically available as a collection of scene of fixed dimension always similar each other when covering the same terrestrial area.