There's been a lot of discussion lately about the exact definitions of varous kinds of orbit: what is the difference between Low Earth Orbit (LEO) and Medium Earth Orbit (MEO)? There's no right answer, since these names are arbitrary. I have my own definitions, which I give below. The boundaries I use are motivated by the physical boundaries in the atmosphere and by historical practice. My proposed definitions: (1) Atmospheric (ATM): suborbital trajectory with apogee less than 80 km (mean height of the mesopause, and same as old USAF definition of 50 miles for astronaut wings) (2) Suborbital spaceflight (SO): suborbital trajectory with apogee more than 80 km and less than lunar distance (3) Transatmospheric orbit (TAO): orbital flight with perigee less than 80 km but more than zero. Potentially used by aerobraking missions and transatmospheric vehicles, also in some temporary phases of orbital flight (e.g. STS pre OMS-2, some failures when no apogee restart) (4) LEO: Low Earth Orbit. Orbits with perigee above 80 km and apogee less than L km. It's not clear what the value of L should be. A histogram of apogee heights for objects currently in orbit shows a big peak from 100 km to about 2500 km, followed by an almost empty region, followed by a small peak at 19000 km (GLONASS and GPS) and another peak at 36000 km (GEO). Why are there so few satellites in the 3000 - 19000 km range? It's because of the radiation belts. Of course polar orbit satellites pass through the radiation belts even at low altitude (the magnetosphere dips into the auroral circle). But at 3000 km and up you pass through the belts at all latitudes. What is the lower level of the radiation belts? I'm still researching this. However, if you look at the apogee histogram in more detail, you see that the lower orbit satellites have two broad peaks: one from 300 km to 1300km peaking at 800-1000 km; and another at 1300-2200 km peaking at 1500 km. This analysis is compromised by the fact that the histogram may be dominated by debris objects from a small number of explosions; it would be better to plot payloads only. Redoing the analysis with only international designations "A" and "B" (e.g. 1997-04B, but not 1997-04BC) gives a similar result but with narrower peaks. In particular, there are very few payloads or rocket stages with apogees in the 1100 to 1400 km or 1600 to 2000 km ranges. I therefore suggest that the LEO/MEO boundary value L should be set at either: apogee 1000 km, a round number definition which would exclude the large number of satellites in the 1000-1100 km range including Parus/Tsikada and Transit navsats. I think 1000 km is a little too low to exclude. apogee 1100 km, a strict definition of LEO apogee 1600 km, a definition including Globalstar and Strela/Gonets and older ESSA/NOAA polar satellites apogee 2000 km, a safe 'round number' definition including all LEO payloads and debris objects. period 120 minutes ( 2 hours ). Another 'round number'. This has an average height of 1680 km and a maximum apogee of 3280 km. With the 2000 km or 2 hr definitions, MEO (Medium Earth Orbit) would be the relatively unpopulated region between LEO and the geosynch corridor, which contains the Glonass and GPS satellites and the old Midas early warning sats, and not much else. I have decided to use the 2 hr definition, but I suspect that the industry may end up using something toward the lower end, say the 1100 km definition. The next boundary of interest is betwee MEO and the 'geosynchronous corridor'. To study the geosynchronous corridor, it's most helpful to work in orbital period and consider drift rates. For a pure equatorial orbit, non-Keplerian perturbations introduce drifts of order 0.05 degrees per day. These dominate Keplerian drift in longitude if the period is roughly between 23h 55.5m and 23h 56.5 min. I call this 'geostationary orbit'. Satellites which are still operational but are being moved from one slot to another usually are drifting at between 0.1 and 10 degrees per day. I find the 10 degree per day drift rate one convenient boundary, corresponding to periods from 23h 17m to 24h 37m (that's what I used to use in my geo.log file). An alternative criterion is to make a period cut from 23h to 25h: 1 hour either side of the geosynch period. I now use personal definitions as follows: Period (hh:mm) Inc (deg) Ecc GEO/S Stationary 23:55.5 - 23:56.5 0.0 - 2.0 0.00 - 0.01 GEO/I Inclined GEO 23:55.5 - 23.56.5 0.0 - 20.0 0.00 - 0.05 GEO/T Synchronous 23:55.5 - 23.56.5 0 - 180 0.00 - 0.85 GEO/D Drift GEO 23:00 - 25:00 0.0 - 2.0 0.00 - 0.05 GEO/ID Inc. Drift GEO 23:00 - 25:00 0.0 - 20.0 0.00 - 0.05 GEO/NS Near-Sync 23:00 - 25:00 0 - 180 0.00 - 0.85 Rather than High Earth Orbit (too easily confused with Highly Elliptical Orbit) I use Deep Space Orbit (DSO), for anything circular above GEO, and Deep Eccentric Orbit (DHEO) for elliptical deep orbits. I then have (with A = apogee, P = perigee, T = period) ATM Atmospheric A < 80 SO Suborbital A >= 80, P < 0 TAO Trans-Atm A >= 80, P = 0 - 80 LEO Low T= 1:26 - 2:00 MEO Medium T= 2:00 - 23:00, e < 0.5 HEO Highly Ellip T= 4:03 - 23:00, e > 0.5 (implies A > 13000) GEO Near-Synch T=23:00 - 25:00 DSO Deep Space T>25:00, e < 0.5 DHEO Deep Eccentric T>25:00, e > 0.5 HCO Heliocentric PCO Planetocentric SSE Solar System Escape subcategories of HEO: HEO/M: Molniya orbit T = 11:30 - 12:30, i = 62.0 - 64.0, e = 0.50 - 0.77 subcategories of LEO: SSO: Sun Synchronous orbit Really T = 3:47 * ( - cos i )** (3/7) +/- 0:10, i = 97.0 - 103.0 The equation relating T and cos i gives a plane precession of 360 deg per year. It's probably good enough to use a less strict but simpler definition of SSO: LEO/S Sun Synch T = 1:26 - 2:00, i = 95.0 - 104.0 One might also usefully define LEO/R Retrograde: T = 1:26 - 2:00, i = 104.0 - 180.0 LEO/P Polar: T = 1:26 - 2:00, i = 85.0 - 95.0 LEO/E Equatorial T = 1:26 - 2:00, i = 0.0 - 20.0